(Sticky Note comment CD1 11/25/2011 9:27:07 AM Unmarked set by CD1) (Typewriter comment SOLUTION MANUAL)

Chapter 1 1.1 (a)One dimensional, multichannel, discrete time, and digital. (b) (c) (d) (e) 1.2 Multi dimensional, single channel, continuous-time, analog. One dimensional, single channel, continuous-time, analog. One dimensional, single channel, continuous-time, analog. One dimensional, multichannel, discrete-time, digital.

0.01 1 (a)f == . periodic with Np = 200. 2 200 30 11 (b)f = ( )= . periodic with Np = 7. 10527 3 3 (c)f = = . periodic with Np = 2. 2 2 3 (d)f = . non-periodic. 2 62 1 31 (e)f = (

)= . periodic with Np = 10. 10 210 1.3 2 (a)Periodic with period Tp = 5 . 5 (b)f = . non-periodic. 2 1 (c)f = . non-periodic. 12 (d)cos(n)is non-periodic; cos(n )isperiodic;Their productis non-periodic. 88 (e)cos(n )is periodic with period Np=4 2 sin(n )is periodic with period Np=16 8

cos(n + )is periodic with period Np=8 43 Therefore, x(n) is periodic with period Np=16. (16 is the least common multiple of 4,8,16). 1.4 2k k (a)w = implies that f =

. Let NN a = GCD of(k,N), i.e., k = k ' a,N = N ' a. Then, k ' f = , which implies that ' NN ' N = . a 3 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) N =7 k = 01234567 GCD(k,N) = 71111117 Np = 17777771 (c) N = 16 k = 0123456789101112 ... 16 GCD(k,N) = 16121412181214 ... 16 Np = 168164168162168164 ... 1 1.5 (a)Refer to fig 1.5-1 (b) -3 -2 -1 0 1 2 3---> xa(t) 0 5 1015202530 ---> t (ms) Figure 1.5-1: x(n)= xa(nT) = xa(n/Fs) =3sin(pn/3) . 1 p f = ( ) 2p31 = ,Np =6 6 4 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 10 20 t (ms) 3 -3 0 10 20 t (ms) 3 -3 Figure 1.5-2: (c)Refer to fig 1.5-2 3v 3v 3v 3v x(n)=0, ,0,,,Np =6. , 2 2 2 2 (d) Yes. 100p x(1)=3 =3sin(). Fs =200 samples/sec. Fs 1.6 (a) x(n)= Acos(2pF0n/Fs + .) = Acos(2p(T/Tp)n + .) But T/Tp = f . x(n)is periodic if f is rational. (b) Ifx(n) is periodic, then f=k/N where N is the period. Then, kTp Td =(

T)= k( )T = kTp. fT Thus, it takes k periods(kTp)ofthe analog signalto make1period(Td)of the discret e signal. (c)Td = kTp . NT = kTp . f = k/N = T/Tp . f is rational . x(n) is periodic. 1.7 (a)Fmax =10kHz . Fs = 2Fmax =20kHz. (b) ForFs =8kHz,Ffold = Fs/2 =4kHz . 5kHz will alias to 3kHz. (c) F=9kHz willalias to 1kHz. 1.8 (a)Fmax =100kHz,Fs = 2Fmax =200Hz. (b)= Fs =125Hz. Ffold 2 5 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1.9 (a)Fmax =360Hz,FN =2Fmax =720Hz. (b)= Fs =300Hz. Ffold 2 (c) x(n)= xa(nT) = xa(n/Fs) = sin(480pn/600)+3sin(720pn/600) x(n)= sin(4pn/5)3sin(4pn/5) = -2sin(4pn/5). Therefore, w =4p/5. (d)ya(t)= x(Fst)= -2sin(480pt). 1.10 (a) Number of bits/sample = log21024 =10. [10,000 bits/sec] Fs = [10 bits/sample] = 1000 samples/sec. = 500Hz. Ffold (b) 1800p Fmax = 2p = 900Hz FN =2Fmax =1800Hz. (c) 600p 1 f1 =( ) 2pFs

=0.3; 1800p 1 f2 =( ) 2pFs =0.9; But f2 =0.9 > 0.5 . f2 =0.1. Hence, x(n) =3cos[(2p)(0.3)n]+2cos[(2p)(0.1)n] xmax-xmin 5-(-5) 10 (d). = == m-1 1023 1023. 1.11 x(n)= xa(nT) 100pn 250pn . =3cos +2sin 200 200 6 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

pn 3pn . =3cos 2sin 24 1 ' ) ' T = . ya(t)= x(t/T 1000 p1000t3p1000t. =3cos 2sin 24 ya(t) =3cos(500pt)2sin(750pt) 1.12 (a)For Fs =300Hz, pn . pn . x(n) =3cos +10sin(pn)cos 6 3 pn pn . =3cos 3cos 63 (b)xr(t)=3cos(10000pt/6)cos(10000pt/3) 1.13

(a) Range = =12.7. xmax xmin range m = 1+ . = 127+1 =128 . log2(128) = 7 bits. 1+ 127 (b)m = =636 . log2(636). 10 bit A/D. 0.02 1.14 samples bits R = (20) (8 ) sec samplebits = 160 sec Fs = =10Hz. Ffold 2 1volt Resolution = 28 1 =0.004. 1.15 (a)Refer to fig 1.15-1. With a sampling frequency of 5kHz, the maximum frequency that can be represented is 2.5kHz. Therefore, a frequency of 4.5kHz is aliased to 500Hz and

the frequency of 3kHz is aliased to 2kHz. 7 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Fs = 5KHz, F0=500Hz Fs = 5KHz, F0=2000Hz Fs = 5KHz, F0=2000Hz 1 1 0.5 0.5 0 0 -0.5 -0.5 -1 -1 0 50 100 0 50 100 Fs = 5KHz, F0=3000Hz Fs = 5KHz, F0=4500Hz 1 1 0.5 0.5 0 0 -0.5 -0.5 -1 -1 0 50 100 0 50 100 Figure 1.15-1: (b) Refer to fig 1.15-2. y(n) is a sinusoidal signal. By taking the even numbere d samples, the samplingfrequencyisreducedto halfi.e.,25kHzwhichisstillgreaterthanthenyquistrate . The frequency of the downsampled signal is 2kHz. 1.16 (a)for levels = 64, using truncation refer to fig 1.16-1.

for levels = 128, using truncation refer to fig 1.16-2. for levels = 256, using truncation refer to fig 1.16-3. 8 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

F0 = 2KHz, Fs=50kHz 0 10 20 30 40 50 60 70 80 90 100 -1 -0.5 0 0.5 1 F0 = 2KHz, Fs=25kHz 0 5 101520253035404550 Figure 1.15-2: levels = 64, using truncation, SQNR = 31.3341dB 1 1 0.5 0.5 -1 -0.5 0 0.5 1 --> xq(n) -0.5 -0.5 -1 -1 0 50 100 150 200 0 50 100 150 200 --> n --> n 0 -0.01 -0.02 -0.03 -0.04 0 50 100 150 200 --> n Figure 1.16-1: 9 0 0 --> x(n)

--> e(n) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

levels = 128, using truncation, SQNR = 37.359dB --> e(n) -1 -0.5 0 0.5 1 --> x(n) -1 -0.5 0 0.5 1 --> xq(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 0 -0.005 -0.01 -0.015 -0.02 0 50 100 150 200 --> n Figure 1.16-2: levels = 256, using truncation, SQNR=43.7739dB -1 -0.5 0 0.5 1 --> x(n) -1 -0.5 0 0.5 1 --> xq(n) --> e(n) -4 -6 -8 0 50 100 150 200 --> n Figure 1.16-3: 10 0 50 100 150 200 0 50 100 150 200

--> n --> n x 10-3 0 -2 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) for levels =64, using rounding refer to fig 1.16-4. for levels = 128, using rounding refer to fig 1.16-5. for levels = 256, using rounding refer to fig 1.16-6. -1 -0.5 0 0.5 1 levels = 64, using rounding, SQNR=32.754dB --> x(n) -1 -0.5 0 0.5 1 --> xq(n) --> e(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 0.04 0.02 0 -0.02 -0.04 0 50 100 150 200 --> n Figure 1.16-4: 11 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

levels = 128, using rounding, SQNR=39.2008dB --> e(n) --> e(n) -1 -0.5 0 0.5 1 --> x(n) -1 -0.5 0 0.5 1 --> xq(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 0.02 0.01 0 -0.01 -0.02 0 50 100 150 200 --> n Figure 1.16-5: levels = 256, using rounding, SQNR=44.0353dB -1 -0.5 0 0.5 1 --> x(n) -1 -0.5 0 0.5 1 --> xq(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 0.01 0.005 0 -0.005 -0.01 0 50 100 150 200

--> n Figure 1.16-6: 12 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c)The sqnr withroundingisgreater than with truncation. But the sqnrimproves as the number of quantization levels are increased. (d) levels 64 128 256 theoretical sqnr 43.9000 49.9200 55.9400 sqnr with truncation 31.3341 37.359 43.7739 sqnr with rounding 32.754 39.2008 44.0353 The theoretical sqnr is given in the table above. It can be seen that theoretica l sqnr is much higher than those obtained by simulations. The decrease in the sqnr is because o f the truncation and rounding. 13 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 2 2.1 (a) 1 2 x(n)= . . . 0, 3, 3,1 . ,1,1,1,0,. . . . Refer to fig 2.1-1. (b) After folding s(n) we have 1111 -3 -2 -1 0 1 2 3 4 1/3 2/3 Figure 2.1-1: 21 x(-n)= ... 0,1,1,1,1, , ,0,... . . 33After delaying the folded signal by 4 samples, we have 21 x(-n +4) = ... 0,0,1,1,1,1, , ,0,... . . 33On the other hand, if we delay x(n) by 4 samples we have 12 x(n 4) = ... 0,0, , ,1,1,1,1,0,... . . 33Now, if we fold x(n 4) we have 21 x(-n 4) = ... 0,1,1,1,1, ,

,0,0,... 33. 15 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) 21 x(-n +4) = ... 0,1,1,1,1, , ,0,... . 33 (d) To obtain x(-n + k), first we fold x(n). This yields x(-n). Then, we shift x (-n) by k samples to the right if k> 0, or k samples to the left if k< 0. (e) Yes. 12 x(n)= d(n 2)+ d(n +1)+ u(n)u(n 4) 33 2.2 11 x(n)= ... 0,1,1,1,1, , ,0,... . 22 (a) 11 x(n 2) = ... 0,0,1,1,1,1, , ,0,... . 22 (b) .. . 11 . x(4n)= ... 0,

, ,1,1,1,1,0,... 22 .. . (see 2.1(d)) (c) 11 x(n +2) = ... 0,1,1,1,1, , ,0,... . 22 (d) x(n)u(2n)= ... 0,1,1,1,1,0,0,... . (e) x(n 1)d(n 3) = ... 0,0,1,0,... . (f) x(n 2)= {... 0,x(4),x(1),x(0),x(1),x(4),0,...} 11 = ... 0, ,1,1,1, ,0,... 2. 2 (g) x(n)+x(-n) xe(n) = 2 , . 1 1 . x(-n) = . . . 0, 2, 2,1,1,1

. ,1,0,. . . . 1 1 1 1 1 1 . = . . . 0, 4, 4, 2,1,1,1, 2, 4, 4,0,. . . 16 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(h) xo(n) = x(n)x(-n) 2 . 1 1 1 1 1 1 . = . . . 0,4,4,2,0,0,0, 2, 4, 4,0,. . . 2.3 (a) . 0, n< 0 . u(n)u(n 1) = d(n)= 1,n =0 . 0, n> 0 (b) n . 0, n< 0 d(k)= u(n)= 1,n = 0 k=-8 8 . 0, n< 0 d(n k)= 1,n = 0 k=0

2.4 Let 1 xe(n)= [x(n)+x(-n)], 2 1 xo(n)= [x(n)x(-n)]. 2 Since xe(-n)= xe(n) and xo(-n)= -xo(n), it follows that x(n)= xe(n)+xo(n). The decomposition is unique. For x(n)=2,3,4,5,6 , . we have xe(n)=4,4,4,4,4 . and xo(n)= -2,-1,0,1,2 . . 17 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.5 First, we prove that 8 n=-8 xe(n)xo(n)= 0 8 . 8 . xe(n)xo(n)= xe(-m)xo(-m) n=-8 m=-8 8 = xe(m)xo(m) m=-8 8 = xe(n)xo(n) n=-8 8 = = n=-8 xe(n)xo(n) 0 Then, 8 8 . x 2(n) = 2 [xe(n)+xo(n)] n=-8 n=-8 888 (n)+ x 2 o

(n)+ 2xe(n)xo(n) 2 = x e n=-8 n=-8 n=-8 = Ee + Eo 2.6 (a)No, the system is time variant. Proof: If x(n). y(n) = x(n 2) x(n k). y1(n) = x[(n k)2] = x(n 2 + k2 2nk) 6= y(n k) (b) (1) x(n)=0,1,1,1,1,0,... . (2) y(n)= x(n 2)= . . . , 0,1,1 . ,1,0,. . . (3) . y(n 2) = ..., 0,0,1,1,1,0,... . 18 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(4) x(n 2) = ..., 0,0,1,1,1,1,0,... . (5) y2(n)= T [x(n 2)] = ..., 0,1,0,0,0,1,0,... . (6) y2(n)6y(n 2) . system is time variant. = (c)(1) x(n)=1,1,1,1 . (2) y(n)=1,0,0,0,0,-1 . (3) y(n 2) =0,0,1,0,0,0,0,-1 . (4) x(n 2) =0,0,1,1,1,1,1 . (5) y2(n)=0,0,1,0,0,0,0,-1 . (6) y2(n)= y(n 2). The system is time invariant, but this example alone does not constitute a proof . (d)(1) y(n)= nx(n), x(n)= ..., 0,1,1,1,1,0,... . (2)

y(n)= ..., 0,1,2,3,... . (3) y(n 2) = ..., 0,0,0,1,2,3,... . (4) x(n 2) = ..., 0,0,0,1,1,1,1,... . 19 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(5) y2(n)= T [x(n 2)] = {..., 0,0,2,3,4,5,...} (6) y2(n)6y(n 2) . the system is time variant. = 2.7 (a)Static, nonlinear, time invariant, causal, stable. (b) Dynamic, linear, time invariant, noncausaland unstable. The latter is easily proved. For the bounded input x(k)= u(k), the outputbecomes n+1 . 0, n< -1 y(n)= u(k)= n +2,n =-1 k=-8 since y(n).8 as n .8, the system is unstable. (c) Static, linear, timevariant, causal, stable. (d) Dynamic, linear, time invariant, noncausal, stable. (e) Static, nonlinear, time invariant, causal, stable. (f)Static, nonlinear, time invariant, causal, stable. (g)Static, nonlinear, time invariant, causal, stable. (h) Static, linear, time invariant, causal, stable. (i)Dynamic, linear, time variant, noncausal, unstable. Note that the bounded inp ut x(n)= u(n)produces an unbounded output. (j)Dynamic, linear, time variant, noncausal, stable. (k)Static, nonlinear, time invariant, causal, stable. (l)Dynamic, linear, time invariant, noncausal, stable. (m)Static, nonlinear, time invariant, causal, stable. (n) Static, linear, time invariant, causal, stable. 2.8 (a)True. If v1(n)= T1[x1(n)]and v2(n)= T1[x2(n)], then a1x1(n)+a2x2(n) yields a1v1(n)+a2v2(n) by the linearity property of T1. Similarly, if y1(n)= T2[v1(n)]and y2(n)= T2[v2(n)],

then 1v1(n)+2v2(n). y(n)= 1y1(n)+2y2(n) by the linearity property of T2. Since v1(n)= T1[x1(n)]and 20 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

v2(n)= T2[x2(n)], it follows that A1x1(n)+A2x2(n) yields the output A1T [x1(n)]+A2T [x2(n)], where T = T1T2. Hence T is linear. (b) True. For T1, if x(n). v(n)and x(n k). v(n k), For T2, if v(n). y(n) andv(n k). y(n k). Hence, For T1T2, if x(n). y(n)and x(n k). y(n k) Therefore, T = T1T2 is time invariant. (c)True. T1 is causal. v(n)dependsonlyonx(k)for k = n. T2 is causal . y(n)depends onlyon v(k)for k = n. Therefore, y(n)depends onlyon x(k)for k = n. Hence, T is causal. (d)True. Combine(a) and(b). (e) True. This follows from h1(n)* h2(n)= h2(n)* h1(n) (f)False. For example, consider T1 : y(n)= nx(n)and T2 : y(n)= nx(n +1). Then, T2[T1[d(n)]] = T2(0)=0.

T1[T2[d(n)]] = T1[d(n +1)] = -d(n +1) 60. = (g)False. For example, consider T1 : y(n)= x(n)+b and T2 : y(n)= x(n)b, where b 60. = Then, T [x(n)]= T2[T1[x(n)]]= T2[x(n)+b]= x(n). Hence T is linear. (h) True. T1 is stable . v(n)is bounded if x(n)is bounded. T2 is stable . y(n)is bounded if v(n)is bounded . 21 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Hence, y(n) is bounded if x(n) is bounded .T = T1T2 is stable. (i)Inverse of(c). T1 and for T2 are noncausal .T is noncausal. Example: T1 : y(n)= x(n +1) and T2 : y(n)= x(n 2) .T : y(n)= x(n 1), whichis causal. Hence,theinverseof(c)isfalse. Inverse of(h): T1 and/or T2 is unstable, implies T is unstable. Example: x(n) T1 : y(n)= e, stable and T2 : y(n)= ln[x(n)], which is unstable. But T : y(n)= x(n), whichis stable. Hence,theinverse of(h)isfalse. 2.9 (a) n y(n)= h(k)x(n k),x(n)=0,n< 0 k=-8 n+Nn+N y(n + N)= h(k)x(n + N k)= h(k)x(n k) k=-8 k=-8 nn+N = h(k)x(n k)+ h(k)x(n k) k=-8 k=n+1 n+N = y(n)+ h(k)x(n k) k=n+1 For a BIBO system, limn.8|h(n)| =0. Therefore, n+N limn.8 h(k)x(n k)=0 and

k=n+1 limn.8y(n + N)= y(N). (b) Letx(n)= xo(n)+au(n), where a is a constant and xo(n)is a bounded signal with lim xo(n)=0. n.8 Then, 88 y(n)= ah(k)u(n k)+ h(k)xo(n k) k=0 k=0 n = ah(k)+yo(n) k=0 22 clearly, . x(n)< 8. . y(n)< 8 (from(c) below)Hence, nono limn.8|yo(n)| =0. 22 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

and, thus, limn.8y(n)= a Pn h(k)= constant. k=0 (c) y(n) = . h(k)x(n k) k 8 8 . #2 . y 2(n) = . . h(k)x(n k) -8 -8 k = . . h(k)h(l) . x(n k)x(n l) k l n But x(n k)x(n l)= x 2(n)= Ex. nn Therefore, y 2(n)= Ex |h(k)||h(l)|. n kl For a BIBO stable system, |h(k)| <M. k Hence, Ey = M2Ex, so that Ey < 0 if Ex < 0. 2.10 The system is nonlinear. This is evident from observation of the pairs x3(n).

y3(n)and x2(n). y2(n). If the system were linear, y2(n)would be of the form y2(n)= {3,6,3} because the system is time-invariant. However, this is not the case. 2.11 since x1(n)+x2(n)= d(n) and the system is linear, the impulse response of the system is y1(n)+y2(n)=0,3,-1,2,1 . . If the system were time invariant, the response to x3(n)would be 3,2,1,3,1 . . But this is not the case. 23 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.12 (a)Any weighted linear combination of the signals xi(n),i =1,2,...,N. (b) Anyxi(n k), where k is anyinteger and i =1,2,...,N. 2.13 A system is BIBO stable if and only if a bounded input produces a bounded output . y(n)= h(k)x(n k) k |y(n)|= |h(k)||x(n k)| k = Mx |h(k)| k where |x(n k)|= Mx. Therefore, |y(n)| < 8 for all n, if and only if |h(k)| < 8. k 2.14 (a)A system is causal . theoutputbecomes nonzeroaftertheinputbecomes non-zero. Hence, x(n)=0for n< no . y(n)=0for n< no. (b) n y(n)= h(k)x(n k), where x(n)=0for n< 0. -8 If h(k)=0for k< 0, then

n y(n)= h(k)x(n k), and hence, y(n)=0for n< 0. 0 On the other hand, if y(n)=0for n< 0, then n h(k)x(n k). h(k)=0,k < 0. -8 2.15 (a) N n For a =1, N M +1 a = n=M N n MM+1 N for a 6=1, + + ... + a = a a a n=M N a n MM+1 a M+1 N+1 = a + a + ... + a N -

a N a (1a) n=M N+1 = a M a 24 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) ForM =0,|a| < 1, and N .8, 8 . 1 n a = ,|a| < 1. 1a n=0 2.16 (a) y(n)= h(k)x(n k) k 8 y(n)= h(k)x(n k)= h(k) x(n k) n nk kn=-8 . !. . = h(k) x(n) kn (b)(1) y(n)= h(n)* x(n)= {1,3,7,7,7,6,4} y(n)=35,h(k)=5,x(k)=7 n kk (2) y(n)= {1,4,2,-4,1} y(n)=4,h(k)=2,x(k)=2 nkk (3) 113 5

y(n)=0, ,, ,-2,0,,-2 2222 y(n)= -5,h(n)=2.5,x(n)= -2 nnn (4) y(n)= {1,2,3,4,5} y(n)=15,h(n)=1,x(n)=15 n nn (5) y(n)= {0,0,1,-1,2,2,1,3} y(n)=8,h(n)=4,x(n)=2 nnn (6) y(n)= {0,0,1,-1,2,2,1,3} y(n)=8,h(n)=2,x(n)=4 nnn (7) y(n)= {0,1,4,-4,-5,-1,3} 25 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. y(n)= -2, . h(n)= -1, . x(n)= 2 n n n (8) y(n)= u(n)+u(n 1)+ 2u(n 2) . y(n)= 8, . h(n)= 8, . x(n)= 4 n n n (9) y(n)= {1,-1,-5,2,3,-5,1,4} . y(n)= 0, . h(n)= 0, . x(n)= 4 n n n (10) y(n)= {1,4,4,4,10,4,4,4,1} . y(n)= 36, . h(n)= 6, . x(n)= 6 n n n (11) y(n)= [2( 1 2)n ( 1 4)n]u(n) . n y(n)= 8 3, . n h(n)= 4 3, . n x(n)= 2 2.17 (a) x(n) h(n) = = . 1 . ,1,1,1 . . 6 .

,5,4,3,2,1 . y(n) = n . k=0 x(k)h(n k) y(0) = x(0)h(0)= 6, y(1) = x(0)h(1)+x(1)h(0)= 11 y(2) = x(0)h(2)+x(1)h(1)+x(2)h(0)= 15 y(3) = x(0)h(3)+x(1)h(2)+x(2)h(1)+x(3)h(0)= 18 y(4) = x(0)h(4)+x(1)h(3)+x(2)h(2)+x(3)h(1)+x(4)h(0)= 14 y(5) = x(0)h(5)+x(1)h(4)+x(2)h(3)+x(3)h(2)+x(4)h(1)+x(5)h(0)= 10 y(6) = x(1)h(5)+x(2)h(4)+x(3)h(2)= 6 y(7) = x(2)h(5)+x(3)h(4)= 3 y(8) = x(3)h(5)= 1 y(n) y(n) = = 0,n = 9 . 6 . ,11,15,18,14,10,6,3,1 . 26 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b)Byfollowingthesame procedureasin(a),weobtain y(n)=6,11,15,18 ,14,10,6,3,1 . (c)Byfollowingthesame procedureasin(a),weobtain y(n)=1,2,2,2,1 . (d)Byfollowingthesame procedureasin(a),weobtain y(n)=1,2,2,2,1 . 2.18 (a) 12 45 x(n) =0, , ,1, , ,2 . 3333 h(n) =1,1,1,1,1 . y(n)= x(n)* h(n) 1 10 20 11 . = ,1,2, ,5, ,6,5, ,2 3. 33 3 (b) 1 x(n)=

n[u(n)u(n 7)], 3 h(n)= u(n +2) u(n 3) y(n)= x(n)* h(n) 1 = n[u(n)u(n 7)]* [u(n +2) u(n 3)] 3 1 = n[u(n)* u(n +2) u(n)* u(n 3)u(n 7)* u(n +2)+ u(n 7)* u(n 3)] 3 1 10 20 y(n)= d(n +1)+ d(n)+2d(n 1)+ d(n 2)+5d(n 3)+ d(n 4)+6d(n 5) 3 33 11 +5d(n -

6)+5d(n 6)+ d(n 7)+d(n 8) 3 2.19 4 y(n)= h(k)x(n k), k=0 x(n)= a-3,a-2,a-1 ,1,a,. . .,a5 . h(n) =1,1,1,1,1 . 27 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

4 y(n) = . x(n k),-3 = n = 9 k=0 = 0, otherwise. Therefore, a-3 y(-3) = , a-3 + a-2 y(-2) = x(-3)+x(-2) = , y(-1) = a-3 + a-2 + a-1 , a-3 + a-2 y(0) = + a-1 +1 a-3 + a-2 y(1) = + a-1 +1+ a, a-3 + a-2 y(2) y(3) y(4) y(5) y(6) y(7) y(8) y(9) a5 2.20 (a)131 x 122 = 15982 (b){1.,3,1}*{1.,2,2} = {1,5,9,8,2} 4 (c)(1+3z + z2)(1+2z +2z2)=1+5z +9z2 +8z3 +2z (d) 1.31x 12.2 = 15.982. (e) These are different ways to perform convolution. 2.21 (a) nn k u(k)bn-k bn y(n)= au(n = = = = = = = = + a-1 +1+ a+ a2 a-1 +1+ a+ a2 + a3 , a4 + a3 + a2 + a+1 a+ a2 + a3 + a4 + a5 , a2 + a3 + a4 + a5 a3 + a4 + a5 , a4 + a5 ,

k)= (ab-1)k k=0 k=0 bn+1n+1 -a u(n),a 6= b b-a bn(n +1)u(n),a = b y(n)= (b) x(n) =1,2,1,1 . h(n) =1,-1,0,0,1,1 . y(n) =1,1,1,0,0,3,3,2,1 . 28 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) x(n) =1,1,1,1,1,0,-1 , . h(n) =1,2,3,2,1 . y(n) =1,3,6,8,9,8,5,1,-2,-2,-1 . (d) x(n) =1,1,1,1,1 , . h ' (n) =0,0,1,1,1,1,1,1 . h(n)= h ' (n)+h ' (n 9), y(n)= y ' (n)+y ' (n 9), where y ' (n) =0,0,1,2,3,4,5,5,4,3,2,1 . 2.22 (a) yi(n)= x(n)* hi(n) y1(n)= x(n)+x(n 1) = {1,5,6,5,8,8,6,7,9,12,12,15,9} , similarly y2(n)= {1,6,11,11,13,16,14,13,15,21,25,28,24,9} y3(n)= {0.5,2.5,3,2.5,4,4,3,3.5,4.5,6,6,7.5,4.5} y4(n)= {0.25,1.5,2.75,2.75,3.25,4,3.5,3.25,3.75,5.25,6.25,7,6,2.25} y5(n)= {0.25,0.5,-1.25,0.75,0.25,-1,0.5,0.25,0,0.25,-0.75,1,-3,-2.25} (b) 1 y3(n)=

y1(n), because 2 1 h3(n)= h1(n) 2 1 y4(n)= y2(n), because 4 1 h4(n)= h2(n) 4 (c)y2(n)and y4(n)are smoother than y1(n), but y4(n)will appear evensmootherbecau seof the smaller scale factor. (d) System 4 results in a smoother output. The negative value of h5(0) is respon sible for the non-smooth characteristics of y5(n) (e) 131 11139. y6(n)= , ,-1, ,1,-1,0, , ,1,, ,22222222 y2(n)is smoother than y6(n). 29

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.23 We can express the unit sample in terms of the unit step function as d(n)= u(n)u(n 1). Then, h(n)= h(n)* d(n) = h(n)* (u(n)u(n 1) = h(n)* u(n)h(n)* u(n 1) = s(n)s(n 1) Using this definition of h(n) y(n)= h(n)* x(n) =(s(n)s(n 1))* x(n) = s(n)* x(n)s(n 1)* x(n) 2.24 If y1(n)= ny1(n 1)+x1(n)and y2(n)= ny2(n 1)+x2(n)then x(n)= ax1(n)+bx2(n) produces the output y(n)= ny(n 1)+x(n), where y(n)= ay1(n)+by2(n). Hence, the system is linear. If the input is x(n 1), we have y(n -

1) =(n 1)y(n 2)+x(n 1). But y(n 1) = ny(n 2)+x(n 1). Hence, the system is time variant. If x(n)= u(n), then|x(n)|= 1. But for this bounded input, the output is y(0)=1,y(1)=1+1 =2,y(2)=2x2+1 =5,... which is unbounded. Hence, the system is unstable. 2.25 (a) d(n)= .(n)a.(n 1) and, d(n k)= .(n k)a.(n k1). Then, 8 x(n)= x(k)d(n k) k=-8 8 = x(k)[.(n k)a.(n k1)] k=-8 30 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

88 x(n)= x(k).(n k)ax(k).(n k1) k=-8 k=-8 88 x(n)= x(k).(n k)ax(k1).(n k) k=-8 k=-8 8 Thus, ck = = k=-8 [x(k)ax(k1)].(n k) x(k)ax(k1) (b) y(n) = T [x(n)] 8 = T [ ck.(n k)] k=-8 8 = ckT [.(n k)] k=-8 8 = ckg(n k) k=-8 (c) h(n)= T [d(n)] = T[.(n)-

a.(n 1)] = g(n)ag(n 1) 2.26 With x(n) = 0, we have 4 y(n 1)+ y(n 1) =0 34 y(-1) = y(-2) 34)2 y(0) =(y(-2) 3y(1) =(4)3 y(-2) 3. . . 4)k+2 y(k) =(y(-2) . zero-input response. 3 2.27 Consider the homogeneous equation: 51 y(n)-

y(n 1)+ y(n 2) =0. 66 The characteristic equation is 51 11 .2 .+ =0.. = , . 66 23 31 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Hence, 11 )n 23The particular solution to x(n)= u(n)is yh(n)= c1( )n + c2( 2n yp(n)= k(2n)u(n). Substitute this solution into the difference equation. Then, we obtain k(2n)u(n)k( 5)(2n-1)u(n 1)+k( 1)(2n-2)u(n 2) =2n u(n) 66 Forn = 2, 5kk 8 4k+ =4 . k = . 36 5 Therefore, the total solution is 8 11 y(n)= yp(n)+yh(n)= (2n)u(n)+c1( )n u(n)+c2( )n u(n). 523To determine c1 and c2, assume that y(-2) = y(-1) =0. Then, y(0)=1 and 5 17 y(1)=

y(0)+2= 66 Thus, 83 + c1 + c2 =1 . c1 + c2 = 55 16 11 17 11 + c2 = . 3c1 +2c2 = c1 + 5236 5 and, therefore, 2 c1 = -1,c2 = . 5 The total solution is 8 1 21 y(n)= (2)n ( )n + ( )n u(n) 5253 32 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.28 Fig. 2.28-1 shows the transient response, yzi(n), for y(-1) = 1 and the steady s tate response, yzs(n). 1 0.8 0.6 0.4 0.2 0 10 8 6 4 2 0 0 5 101520253035404550 Normalized Transient Response 0 5 101520253035404550 Steady State Response Figure 2.28-1: 2.29 h(n)= h1(n)* h2(n) 8 = a k=-8 k[u(k)u(kN)][u(n k)u(n kM)] 88 = a

k u(k)u(n k)k a u(k)u(n kM) k=-8 k=-8 88 k a u(kN)u(n k)+ k u(kN)u(n kM) a k=-8 k=-8 . nn-M nn-M = a k k a k k a

a k=0 k=0 k=Nk=N =0 33 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.30 y(n)3y(n 1)4y(n 2) = x(n)+2x(n 1) The characteristic equation is .2 3.4 =0. Hence, . =4,-1 and yh(n)= c1(n)4n + c2(-1)n . Since 4 is a characteristic root and the excitation is x(n)=4n u(n), we assume a particular solution of the form yp(n)= kn4n u(n). Then kn4n u(n)3k(n 1)4n-1 u(n 1)4k(n 2)4n-2 u(n 2) =4n u(n)+2(4)n-1 u(n 1) . For n =2, k(3212) =42 +8 =24 . k = 6 . 5 The total solution is y(n)= yp(n)+yh(n) 6 . = n4n + c14n + c2(-1)n u(n) 5 To solve for c1 and c2, we assume that y(-1) = y(-2) =0. Then, y(0)=1 and y(1)=3y(0)+4+2=9 Hence, c1 + c2 =1 and 24 +4c1 c2 =9 5

21 4c1 c2 = 5 Therefore, 26 1 c1 = and c2 = 25 25 The total solution is 6 26 1 . 4n y(n)= n4n + (-1)n u(n) 5 25 25 34 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.31 From 2.30, the characteristic values are . =4,-1. Hence yh(n)= c14n + c2(-1)n When x(n)= d(N), we find that y(0)=1and y(1)3y(0)=2 or y(1)=5. Hence, c1 + c2 =1 and4c1 c2 =5 6 This yields, c1 = and c2 = 51 . Therefore, 5 6 1 4n h(n)= (-1)n u(n) 55 2.32 (a)L1 = N1 + M1 and L2 = N2 + M2 (b) Partialoverlap from left: low N1 + M1 high N1 + M2 1 Full overlap: low N1 + M2 high N2 + M1 Partial overlap from right: low N2 + M1 +1 high N2 + M2 (c) x(n) =1,1,1,1,1,1,1 . h(n) =2,2,2,2 .

N1 N2 M1 M2

= -2, =4, = -1, =2,

Partial overlap from left: n = -3 n = -1 L1 = -3 Full overlap: n =0 n =3 Partial overlap from right:n =4 n =6 L2 =6 35 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.33 (a) y(n)0.6y(n 1)+0.08y(n 2) = x(n). The characteristic equation is .2 0.6.+0.08 =0. . =0.2,0.4 Hence, nn 12 yh(n)= c1 + c2 . 55 With x(n)= d(n), the initial conditions are y(0) =1, y(1)0.6y(0) =0 . y(1)=0.6. Hence,c1 + c2 = 1 and 12 =0.6 . c1 = -1,c2 =3. c1 + 55 12 Therefore h(n)= -( )n +2( )n u(n) 55 The step response is n s(n) = . h(n k),n = 0 k=0 = n

. . 2( 2 5)n-k ( 1 5)n-k . k=0 . 1 . 2n+1 . 1 . 1n+1 . = 0.12 ( 5 1 0.16 ( 5 1 u(n) (b) y(n)0.7y(n 1)+0.1y(n 2) =2x(n)x(n 2). The characteristic equation is .2 0.7.+0.1 =0. 11 . = 25 Hence, , nn 11 yh(n)= c1 + c2 . 25 With x(n)= d(n), we have y(0) =2, y(1)0.7y(0) =0 . y(1)=1.4.

Hence,c1 + c2 = 2 and 11 7 =1.4= c1 + 25 5 2 14 c2 = . . c1 + 55 These equations yield 10 4 c1 = ,c2 = . 33 10 1 41 . )n h(n)= ( ( )n u(n) 3235 36 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The step response is n s(n)= h(n k), k=0 nn 10 . 14 . 1 )n-k )n-k = ( ( 3235 k=0 k=0 nn 10 1 . 41 . )n 2k )n 5k 3235 = ( ( k=0 k=0 nn 10 1 11 (2n+1 1)u(n)(5n+1 1)u(n) 32 35 =

( ( 2.34 111 1 h(n) =1, , , , . 24816 y(n) =1,2,2.5,3,3,3,2,1,0 . x(0)h(0) = y(0). x(0)=1 13 x(0)+x(1) = y(1). x(1)= 22 By continuing this process, we obtain 3373 x(n)=1, , , , ,... 2242 2.35 (a)h(n)= h1(n)* [h2(n)h3(n)* h4(n)] (b) h3(n)* h4(n) =(n 1)u(n 2)

h2(n)h3(n)* h4(n) =2u(n)d(n) 11 1 h1(n)= d(n)+ d(n 1)+ d(n 2) 24 2 11 1 . Hence h(n)= d(n)+ d(n 1)+ d(n 2) * [2u(n)d(n)] 24 2 15 5 = d(n)+ d(n 1)+2d(n 2)+ u(n 3) 24 2 (c) x(n) =1,0,0,3,0,-4 . 15 25 13 . y(n)= ,

,2, , ,5,2,0,0,... 24. 42 37 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.36 First, we determine s(n) = u(n)* h(n) 8 s(n)= u(k)h(n k) k=0 n = h(n k) k=0 8 n-k = a k=0 n+1 1 a = ,n = 0 a 1 For x(n)= u(n +5) u(n 10), we have the response n+6 1 n-9 1 aa s(n s(n 10) u(n 10) +5) = u(n +5) -

a 1 a 1 From figure P2.33, y(n)= x(n)* h(n)-

x(n)* h(n 2) n+6 1 n-9 1 aa Hence, y(n)= u(n +5) u(n 10) a 1 a 1 n+4 1 n-11 1 aa u(n +3)+ u(n 12) a 1 a 1 2.37 h(n) =[u(n)u(n M)]/M 8 s(n)= u(k)h(n k) k=-8 n n+1 M , n<M = h(n k)= 1,n = M k=0 2.38 8 8 . n=-8

|h(n)| = . n=0,neven 8 |a|n = n=0 |a|2n Stable if |a| < 1 = 1 1|a|2 38 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.39 n h(n)= au(n). The response to u(n)is 8 y1(n)= u(k)h(n k) k=0 n n-k = a k=0 n n -k = aa k=0 n+1 1a = u(n) 1a Then, y(n)= y1(n)y1(n 10) = 1(1a n+1)u(n)(1a n-9)u(n 10) 1a 2.40 We may use the result in problem 2.36 with a = 2. 1 Thus, 1)n+1 1)n-9 y(n)=21(

u(n)21( u(n 10) 22 2.41 (a) 8 y(n) = . h(k)x(n k) k=-8 = n . ( 1 2)k2n-k k=0 = 2n n . ( 1 4)k k=0 = 2n . 1( 1 4)n+1 . ( 4 3) = 2 3 . 2n+1 ( 1 2)n+1 . u(n) (b) 8 y(n)= h(k)x(n k) k=-8 8 = h(k) k=0

8 . 1)k =( =2,n< 0 2 k=0 39 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 . y(n)= h(k) k=n 8 X 1)k ( = 2 k=n 8 . n-1 1 1 )k )k ( ( = 2 2 k=0 k=0 1(1)n 2 =2() 1 2 1 = 2(

)n ,n = 0. 2 2.42 (a) he(n)= h1(n)* h2(n)* h3(n) =[d(n)d(n 1)]* u(n)* h(n) =[u(n)u(n 1)]* h(n) = d(n)* h(n) = h(n) (b) No. 2.43 (a)x(n)d(n n0)= x(n0). Thus, only the value of x(n)at n = n0 is of interest. x(n)* d(n n0)= x(n n0). Thus, we obtain the shifted version of the sequence x(n). (b) 8 . y(n) = h(k)x(n k) k=-8 = h(n)* x(n) Linearity:x1(n). y1(n) = h(n)* x1(n) x2(n). y2(n) = h(n)* x2(n) Then x(n) = ax1(n)+x2(n). y(n)= h(n)* x(n) y(n) = h(n)* [ax1(n)+x2(n)] = ah(n)* x1(n)+h(n)* x2(n) = ay1(n)+y2(n)

Time Invariance: x(n). y(n) = h(n)* x(n) x(n n0). y1(n) = h(n)* x(n n0) = = k h(k)x(n n0 k) y(n n0) (c)h(n)= d(n n0). 40 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.44 (a)s(n)= -a1s(n 1)a2s(n 2)... aNs(n N)+b0v(n). Refer to fig 2.44-1. 1 (b)v(n)= b0 [s(n)+a1s(n 1)+a2s(n 2)+... + aNs(n N)]. Refer to fig 2.44-2 s(n) v(n) + + + + z-1 z-1 z -1 -a 1 -a 2 -a N b0 Figure 2.44-1: 41 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

s(n) 1/b0 v(n) a1 a2 aN z -1 z -1 z-1 + + + + s(n) 1/b0 v(n) a1 a2 aN z -1 z -1 z-1 + + + + Figure 2.44-2: 2.45 1 y(n)= y(n 1)+x(n)+2x(n 2) 21 y(-2) = y(-3)+x(-2)+2x(-4) =1 213 y(-1) = y(-2)+x(-1)+2x(-3) = 22 1 17 y(0) = -

y(-1)+2x(-2)+x(0)= 24 1 47 y(1) = y(0)+x(1)+2x(-1) = , etc 28 2.46 (a)Refer to fig 2.46-1 (b) Refer to fig 2.46-2 42 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) 1/2 -1/2+ + 2 z-1 z-1 z -1 z -1 z -1 3/2 + y(n) x(n) 1/2 -1/2+ + 2 z-1 z-1 z -1 z -1 z -1 3/2 + y(n) Figure 2.46-1: 2.47 (a) x(n) =1,0,0,... . 1 y(n)= y(n 1)+x(n)+x(n 1) 2y(0) = x(0)=1, 13 y(1) = y(0)+x(1)+x(0)= 22 13 y(2) = y(1)+x(2)+x(1)= . Thus, we obtain

24 3333 3 y(n) =1, , , , , ,... 2481632 1 (b)y(n)= y(n 1)+x(n)+x(n 1) 2 (c) As in part(a), we obtain 5 13 29 61 y(n)=1, , , , ,... 24 816 (d) y(n)= u(n)* h(n) = u(k)h(n k) k 43 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim

itris G. Manolakis. ISBN 0-13-187374-1.

y(n) x(n) z -1 z -1 z -1 z -1 -1 -2 -3 + + + Figure 2.46-2: n = h(n k) k=0 y(0) = h(0)=1 5 y(1) = h(0)+h(1)= 2 13 y(2) = h(0)+h(1)+h(2)= , etc 4 (e) from part(a),h(n)=0for n< 0 . the system is causal. 8 X3 1 1 n=0 |h(n)| = 1+ 2(1+ 2 + 4 + . . .)= 4 . system is stable 2.48 (a) y(n) = ay(n 1)+bx(n) . h(n) = ban u(n)

8 b h(n)= =1 1a n=0 . b =1a. 44 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) n s(n)= h(n k) k=0 n+1 1a = bu(n) 1a b s(8)= =1 1a . b =1a. (c)b =1a in both cases. 2.49 (a) y(n) =0.8y(n 1)+2x(n)+3x(n 1) y(n)0.8y(n 1) =2x(n)+3x(n 1) The characteristic equation is .0.8 =0 . =0.8. yh(n)= c(0.8)n Let us first consider the response of the sytem y(n)0.8y(n 1) = x(n) to x(n)= d(n). Since y(0)= 1, it folows that c = 1. Then, the impulse response o f the original system is h(n) = 2(0.8)n u(n)+3(0.8)n-1 u(n 1) =2d(n)+4.6(0.8)n-1 u(n -

1) (b) The inverse system is characterized bythe difference equation 1 x(n)= -1.5x(n 1)+ y(n)0.4y(n 1) 2 Refer to fig 2.49-1 2.50 y(n)=0.9y(n 1)+x(n)+2x(n 1)+3x(n 2) (a)Forx(n)= d(n), we have y(0) y(1) y(2) y(3) y(4) y(5) 45 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1. =1, =2.9, =5.61, =5.049, =4.544, =4.090,...

y(n) -1.5 -0.4 z -1 0.5 + x(n) + y(n) -1.5 -0.4 z -1 0.5 + x(n) + Figure 2.49-1: (b) s(0) s(1) s(2) s(3) 4 s(4) = y(n)=19.10 0 5 s(5) = y(n)=23.19 0 (c) h(n) =(0.9)n u(n)+2(0.9)n-1 u(n 1)+3(0.9)n-2 u(n 2) = d(n)+2.9d(n 1)+5.61(0.9)n-2 u(n 2) 2.51 (a) 11 y(n)= x(n)+ x(n 3)+y(n 1) 33 for x(n)= d(n), we have 1 = = = = y(0)=1, y(0)+y(1)=3.91 y(0)+y(1)+y(2)=9.51 y(0)+y(1)+y(2)+y(3)=14.56

112222 h(n)= , , , , , , ,... 3333333 (b) 111 y(n)= y(n 1)+ y(n 2)+ x(n 2) 282 with x(n)= d(n), and y(-1) = y(-2) =0, we obtain 113111 15 41 h(n) =0,0, , , , , , , ,... 241681282561024 46 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) y(n) = 1.4y(n 1)0.48y(n 2)+x(n) with x(n) = d(n), and y(-1) = y(-2) = 0, we obtain h(n) = {1,1.4,1.48,1.4,1.2496,1.0774,0.9086,. . .} (d) All three systems are IIR. (e) y(n) The characteristic equation is .2 1.4.+0.48 . yh(n) c1 + c2 0.8c1 +0.6c2 . c1 c2 h(n) 2.52 (a) =1.4y(n 1)0.48y(n 2)+x(n) = 0 Hence =0.8,0.6. and = c1(0.8)n + c2(0.6)nFor = 1 and =1.4 =4, = -3. Therefore = [4(0.8)n 3(0.6)n]u(n) x(n)= d(n). We have, h1(n) = c0d(n)+c1d(n 1)+c2d(n 2) h2(n) = b2d(n)+b1d(n 1)+b0d(n 2) h3(n) = a0d(n)+ (a1 + a0a2)d(n 1)+a1a2d(n 2) (b) The only question is whether

h3(n) ? = h2(n)= h1(n) Let a0 = c0, a1 + a2c0 = c1, a2a1 = c2. Hence c2 + a2c0 c1 = 0 a2 . c0a 2 2 c1a2 + c2 = 0 For c0 60, the quadratic has a real solution if and only if = 2 c1 4c0c2 = 0 2.53 (a) 1 y(n) = 2y(n 1)+x(n)+x(n 1) 1 For y(n)2y(n 1) = d(n), the solution is h(n) = ( 1 2)n u(n)+ ( 1 2)n-1 u(n 1) 47 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

)n-1 (b)h1(n)* [d(n)+d(n 1)] =(1)nu(n)+(1 u(n 1). 22 2.54 (a) convolution: y1(n) =1,3,7,7,7,6,4 . correlation: .1(n) =1,3,7,7,7,6,4 . (b) 131 5 . convolution: y2(n)= ,0, ,-2, ,-6,,-2 2. 222 131 5 . correlation: .1(n)= ,0, ,-2, ,-6,,-2 2. 222 Note that y2(n)= .2(n), because h2(-n)= h2(n)(c) convolution: y3(n) =4,11,20,30,20,11,4 .

correlation: .1(n) =1,4,10,20 ,25,24,16 . (c) convolution: y4(n) =1,4,10,20,25,24,16 . correlation: .4(n) =4,11,20,30 ,20,11,4 . Note that h3(-n)= h4(n +3), hence, .3(n)= y4(n +3) and h4(-n)= h3(n +3), . .4(n)= y3(n +3) 2.55 Obviously, the length of h(n)is 2, i.e. h(n)= {h0,h1} h0 =1 3h0 + h1 =4 . h0 =1,h1 =1 2.56 NM (2.5.6) y(n)= aky(n k)+ bkx(n k) k=1 k=0 48 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N (2.5.9) w(n)= akw(n k)+x(n) k=1 M (2.5.10) y(n)= bkw(n k) k=0 From(2.5.9) we obtain N x(n)= w(n)+ akw(n k) (A) k=1 By substituting(2.5.10) for y(n)and(A) into(2.5.6),weobtainL.H.S =R.H.S. 2.57 y(n)4y(n 1)+4y(n 2) = x(n)x(n 1) The characteristic equation is .2 4.+4 =0 . =2,2. Hence, yh(n)= c12n + c2n2n The particular solution is yp(n)= k(-1)n u(n). Substituting this solution into the difference equation, we obtain k(-1)n u(n)4k(-1)n-1 u(n 1)+4k(-1)n-2 u(n 2) =(-1)n u(n)(-1)n-1 u(n 1) 2 For n =2,k(1+4+4)=2 . k = 9. The total solution is 2 y(n)= c12n + c2n2n +

(-1)n u(n) 9 From the initial condtions, we obtain y(0)=1,y(1)=2. Then, 2 =1 c1 + 9 7 . c1 = , 9 2 2c1 +2c2 =2 9 1 . c2 = , 3 2.58 From problem 2.57, h(n)=[c12n + c2n2n]u(n) With y(0)=1,y(1)=3, we have c1 =1 2c1 +2c2 =3 49 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 . c2 = 2 Thus h(n) = 1 2n n2n + 2 u(n) 2.59 x(n) = x(n)* d(n) = x(n)* [u(n)u(n 1)] = [x(n)x(n 1)]* u(n) 8 =[x(k)x(k1)]u(n k) k=-8 2.60 Let h(n)be the impulse response of the system k . s(k) = h(m) m=-8 . h(k) = s(k)s(k1)

8 y(n)= h(k)x(n k) k=-8 8 = [s(k)s(k1)]x(n k) k=-8 2.61 x(n)= 1,n0 N = n = n0 + N 0, otherwise y(n)= 1, -N = n = N 0, otherwise 8 .xx(l)= x(n)x(n l) n=-8 The range of non-zero values of .xx(l)is determined by n0 N = n = n0 + N n0 N = n l = n0 + N which implies -2N = l = 2N For a given shift l, the number of terms in the summation for which both x(n)and x(n l)are non-zero is 2N +1 -|l|, and the value of each term is 1. Hence,

.xx(l)= 2N +1 -|l|, -2N = l = 2N 0, otherwise 50 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

For .xy(l)we have 2N +1 -|ln0|,n0 2N = l = n0 +2N 0, otherwise .xy(l)= 2.62 (a) 8 .xx(l)= x(n)x(n l) n=-8 .xx(-3) = x(0)x(3)=1 .xx(-2) = x(0)x(2)+x(1)x(3)=3 .xx(-1) = x(0)x(1)+x(1)x(2)+x(2)x(3)=5 3 .xx(0) 2(n) = 7 = x n=0 Also .xx(-l)= .xx(l) Therefore .xx(l) =1,3,5,7,5,3,1 . (b) 8 .yy(l)= y(n)y(n l) n=-8 We obtain .yy(l)= {1,3,5,7,5,3,1}

we observe that y(n)= x(-n +3), which is equivalent to reversing the sequence x( n). This has not changed the autocorrelation sequence. 2.63 8 .xx(l)= x(n)x(n l) n=-8 2N +1 -|l|, -2N = l = 2N 0, = otherwise .xx(0) =2N +1 Therefore, the normalized autocorrelation is 1 .xx(l) =(2N +1 -|l|),-2N = l = 2N 2N +1=0, otherwise 2.64 (a) 8 .xx(l)= x(n)x(n l) n=-8 51 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 =[s(n)+.1s(n k1)+.2s(n k2)]* n=-8 [s(n l)+.1s(n lk1)+.2s(n lk2)] = (1+.12 + .22).ss(l)+.1 [.ss(l+ k1)+.ss(lk1)] +.2 [.ss(l+ k2)+.ss(lk2)] +.1.2 [.ss(l+ k1 k2)+.ss(l+ k2 k1)] (b) .xx(l) has peaks at l =0,k1,k2 and (k1 + k2). Suppose that k1 <k2. Then, we can determine .1 and k1. The problem is to determine .2 and k2 from the other peaks. (c) If.2 =0,thepeaks occur at l =0 and l = k1. Then, it is easy to obtain .1 and k1. 2.65 (a)The shift at which the crosscorrelation is maximum is the amount of delay D. (b) variance= 0.01. Refer to fig 2.65-1. (b) DelayD = 20. Refer to fig 2.65-1. -1 -0.5 0 0.5 1 --> x(n) -1.5 -1 -0.5 0 0.5 1 1.5 --> y(n) --> rxy(l) 0 50 100 150 200 0 50 100 150 200 --> n --> n 15 10 5

0 -5 -20 0 20 --> l Figure 2.65-1: variance = 0.01 (c) variance= 0.1. Delay D = 20. Refer to fig 2.65-2. (d) Variance= 1. delay D = 20. Refer to fig 2.65-3. (e)x(n)= {-1,-1,-1,+1,+1,+1,+1,-1,+1,-1,+1,+1,-1,-1,+1}. Refer to fig 2.65-4. (f)Refer to fig 2.65-5. 52 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 -0.5 0 0.5 1 --> x(n) -1.5 -1 -0.5 0 0.5 1 1.5 --> y(n) -1 -0.5 0 0.5 1 --> x(n) -1.5 -1 -0.5 0 0.5 1 1.5 --> y(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 20 15 10 5 0 -5 -20 0 20 Figure 2.65-2: variance = 0.1 -1 -0.5 0 0.5 1 --> x(n) -3 -2 -1 0 1 2 3 --> y(n)

0 50 100 150 200 0 50 100 150 200 --> n --> n 15 10 -20 0 20 5 0 -5 --> l Figure 2.65-3: variance = 1 53 --> rxy(l) --> rxy(l) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 -0.5 0 0.5 1 --> x(n) -1.5 -1 -0.5 0 0.5 1 --> y(n) --> x(n) -1.5 -1 -0.5 0 0.5 1 --> y(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 20 15 10 5 0 -5 -10 -20 0 20 --> n Figure 2.65-4: -1 -0.5 0 0.5 1 --> x(n) -1.5 -1 -0.5 0 0.5 1 1.5 --> y(n) 0 50 100 150 200 0 50 100 150 200 --> n --> n 20 15

-20 0 20 10 5 0 -5 -10 --> n Figure 2.65-5: 54 --> rxy(l) --> rxy(l) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.66 (a)Refer to fig 2.66-1. (b) Refer to fig 2.66-2. 0 5 10 15 20 25 30 35 40 45 50 -0.5 0 0.5 1 --> h(n) impulse response h(n) of the system --> n Figure 2.66-1: (c) Refer to fig 2.66-3. (d) The step responses in fig 2.66-2 and fig 2.66-3 are similar except for the s teady state value after n=20. 55 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

zero-state step response s(n) 0 5 101520253035404550 --> n Figure 2.66-2: 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 --> s(n) step response 0 5 101520253035404550 --> n Figure 2.66-3: 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 --> s(n) 56 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2.67 Refer to fig 2.67-1. -2 -1 0 1 2 3 4 5 6 7 --> h(n) 0 10 20 30 40 50 60 70 80 90100 --> n Figure 2.67-1: 57 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

58 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 3 (a) X(z)= x(n)z n -n =3z 5 +6+ z -1 4z -2 ROC: 0 < |z| < 8 (b) X(z)= x(n)z n -n 8 X 1 -n )n ( z = 2 n=5 8 1 )n ( = 2z n=5 8 1 -1)m+5

( z = 2 = = m=0 ( z-1 2 )5 1 11 2z-1 ( 1 32) z-5 11 2z-1 ROC: |z| > 1 2 3.2 (a) X(z)= x(n)z n 8 -n = n=0 -n (1+n)z 88 -n -n + = z nz

n=0 n=0 8 1 -n But ROC: |z| > 1 z = 1z-1 n=0 59 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

and 8 -1 z -n nz = (1z-1)2 n=0 ROC: |z| > 1 Therefore, X(z) = = 1z-1 (1z-1)2 + z-1 (1z-1)2 1 (1z-1)2 (b) 8 X(z) =(a n -n -n)z + a n=0 88 -n -n -n n + = az az n=0 n=0

8 1 n -n But ROC: |z| > |a| az = 1az-1 n=0 8 11 -n -n and ROC: |z| > az = (11 z-1)2 a |a| n=0 11 Hence, X(z)= + 1az-1 11z-1 a 2(a + 1 -1 a = )zROC: |z| > max(|a|, 1

) (1az-1)(11z-1) |a| a (c) 8 1 -n )n (z X(z) = 2 n=0 11 = ,|z| > 1 z-12 1+ 2 (d) 8 X(z)= na n=0 8 -n nsinw0nz jw0n e-jw0n e

-n n = na z 2j n=0 1 ae jw0z-1 -jw0z-1 ae = 2j (1aejw0z-1)2 (1ae-jw0z-1)2 -1)3. az-1 (azsinw0 ,|z| > (12acosw0z-1 + a2z-2)2 = a (e) 8 X(z)= na n -n cosw0nz n=0 8 jw0n -jw0n e+ e

-n n = na z 2 n=0 60 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. jw0z-1 -jw0z-1 1 aeae = + 2 (1aejw0z-1)2 (1ae-jw0z-1)2 . -1)3. 2-2 az-1 +(azsinw0 2az = , |z| >a (12acosw0z-1 + a2z-2)2 (f) 8 n -n X(z)= A r cos(w0n + f)z n=0 8 . jw0nj -jw0n-j . n ee+ ee -n = Ar z 2 n=0 . j -j Aee = + 21rejw0z-1 1re-jw0z-1 -1

. cosf rcos(w0 f)z = A, |z| >r 12rcosw0z-1 + r2z-2 (g) 8 . 1 2 1)n-1 -n X(z)= (n + n)( z 23 n=1 8 (1 -1 -1 )3zz . 1)n-1 -13 But n( z == 3(11z-1)2 (11z-1)2 n=1 33 8 -11 -2 . 1 z+ z )n-1 -n 3 2( n z =

3(11 z-1)3 n=1 3 . -1 -11 -2 z 1 zz+ 3 Therefore, X(z)= + 2 (11z-1)2 (11z-1)3 33-1 z1 = , |z| > (11 z-1)33 3 (h) 88 . 1 . 1 -n )n X(z) =( )n z -n ( z 22 n=0 n=10 (1)10-10 z 1 2

= 11 z-1 11 z-1 22-1)10 1(1 z1 2 = , |z| > 11 z-1 2 2 The pole-zero patterns are as follows: (a)Double pole at z =1 and a zero at z =0. 1 11 (b) Poles atz = a and z = . Zeros at z =0 and z = (a + ). a 2a (c) Pole atz = 1 and zero at z =0. 2 (d) Double poles atz = aejw0 and z = ae-jw0 and zeros at z =0, z = a. (e) Double poles atz = aejw0 and z = ae-jw0 and zeros are obtained by solving th e quadratic 23 acosw0z 2 2az + a cosw0 =0. (f)Poles at z = rejw0 and z = ae-jw0 and zeros at z =0, and z = rcos(w0 f)/cosf. 11 (g)Triple pole at z = and zeros at z =0 and z =

.Hence thereisapole-zero cancellationso 33 61 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 that in reality there is only a double pole at z = and a zero at z =0. 3 (h)X(z)has a pole of order 9 at z =0. For nine zeros which we find from the root s of 1 -1)10 1( z =0 2 1)10 z 10 or, equivalently,( =0 2 1 j2n 10 Hence, ,n = 1,2,. . .,k. zn = e 2 1 Note the pole-zero cancellation at z = 2. 3.3 (a) 8 0 X 1

1 -n )n z -n 1 )n X1(z) =( +( z 3 2 n=0 n=-8 8 1 1 ( )n z n 1 =+ 11 z-1 2 3n=0 11 =+ 1, 113z-1 112z 5 6 = (11z-1)(11z)

32 The ROC is 1 < |z| < 2. 3 (b) 88 1)n -n z -n 2n X2(z) =( z 3 = = n=0 n=0 1 11 3z-1 1 12z-1, 5 3z-1 (11 3z-1)(12z-1) The ROC is |z| > 2. (c) 8 X3(z) = x1(n + 4)z n=-8 = z 4X1(z) = 5 6z4 -n (11 z-1)(11 z)

32 The ROC is 1 < |z| < 2. 3 (d) 8 X4(z)= x1(-n)z n=-8 62 -n (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 m x1(m)z = m=-8 -1) = X1(z 5 6 = (11 z)(11 z-1) 32 The ROC is 1 < |z| < 3. 2 (a) 8 -n n(-1)n z X(z)= n=0 8 d -n (-1)n z = -z dz n=0 d 1

= -z z-1 dz 1+ -1 z = ,|z| > 1 (1+z-1)2 (b) 8 . X(z)= n 2 -n z n=0 8 2 d2 -n = z z dz2 n=0 d2 1 2 = z dz2 1z-1 -1 -1 z2z = + (1-

z-1)2 (1z-1)3 z-1(1+z-1) = ,|z| > 1 (1z-1)3 (c) -1 X(z)= -na n -n z n=-8 -1 d -n = -z a(n)z dz n=-8 d 1 = -z dz 1az-1 -1 az = ,|z| < |a| (1az-1)2 (d) 8 p -n

(-1)n cos( n)z X(z) = 3 n=0 63 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Fromformula(9) intable3.3 with a = -1, -1 1+zcos 3 X(z)= z-2 1+2z-1cos 3 + 1 -1 1+ z 2 = , ROC: |z| > 1 1+z-1 + z-2 (e) 8 X(z)= -n (-1)n z n=0 1 = , |z| > 1 z-1 1+ (f) x(n) X(z) = = 1 . ,0,-1,0,1,-1 1z -2 + z -4 -

z -5 , z 6= 0 3.5 Right-sided sequence :xr(n) = 0,n < n0 -1 8 Xr(z)= xr(n)z -n + xr(n)z -n n=n0 n=0 The term P-1 xr(n)z-n converges for all z except z = 8. n=n0 -n The term P8 (n)zconverges for all |z| >r0 where some r0. Hence Xr(z)converges for n=0 xr r0 < |z| < 8 when n0 < 0 and |z| >r0 for n0 > 0 Left-sided sequence :xl(n) =0,n> n0 0 n0 Xl(z)= xl(n)z -n + xl(n)z -n n=-8 n=1 The first term converges for some |z| <rl. The second term converges for all z, except z = 0. Hence, Xl(z)converges for 0 < |z| <rl when n0 > 0, and for |z| <rl when n0 < 0. Finite-Duration Two-sided sequence :x(n) =0,n> n0 and n < n1, where n0 >n1 n0 X(z)= x(n)z n=n1 -1 -n n=n0 =

x(n)z -n + x(n)z -n n=n1 n=0 The first term converges everywhere except z = 8. The second term converges everywhere except z =0. Therefore, X(z)converges for 0 < |z| < 8. 64 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

n y(n)= x(k) k=-8 . y(n)y(n 1) = x(n) -1 Hence,Y(z)Y(z)z = X(z) X(z) Y(z)= 1z-1 (1)n ,n = 0 3 x1(n)= (12)-n , n< 0 8-1 1 1 )n z -n )-n -n z X1(z) =( +( 3 2 n=0 n=-8 11 =+ 1

11 z-1 11 z 325 6 = (11 z-1)(11 z) 32 8 X 1 -n )n z X2(z) =( 2 n=0 11 = , < |z| < 2 11 z-12 2 10 -4 -2 33 Then,Y(z)= ++ 11 z-1 11 z-1 12z-1 32 10

-2(1)n + (1)n ,n = 0 332 Hence,y(n)= 4 (2)n , n< 0 3 (a) n y(n)= x(k) k=-8 8 = x(k)u(n k) k=-8 = x(n)* u(n) Y(z)= X(z)U(z) X(z) = 1z-1 (b) 8 u(n)* u(n)= u(k)u(n k) k=-8 65 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

n = u(k)=(n +1)u(n) k=-8 Hence,x(n)= u(n)* u(n) 1 andX(z)= ,|z| > 1 (1z-1)2 3.9 y(n)= x(n)ejw0n . From the scaling theorem, we have Y(z)= X(e-jw0z). Thus, the p oles and zeros are phase rotated by an angle w0. 3.10 1 x(n)= [u(n)+(-1)n u(n)] 2 11 ( -1 + -1 ) 1-z1+z X+(z)= 2 From the final value theorem x(8) = lim (z 1)X+(z) z.1 z(z 1) = lim (z +) z.1z +1 1 =

2 3.11 (a) 4 1+2z X(z)= 12z-1 + z-2 -3 = 1+4z -1 +7z -2 +10z + ... Therefore,x(n) =1,4,7,10,..., 3n +1,... . (b) 3 X(z) =2z +5z 2 +8z + ... Therefore,x(n)= ..., -(3n +1),..., 11,8,5,2,0 . 3.12 1 X(z)= (12z-1)(1z-1)2 A BCz-1 = ++ (12z-1) (1z-1) (1z-1)2 A =4,B = -3,C = -1 Hence,x(n) = [4(2)n 3n]u(n) 66 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

3.13 (a) x(n),n even 2 x1(n)= 0,n odd 8 -n X1(z)= x1(n)z n=-8 8 n -n x( )z = 2 n=-8 8 = k=-8 = X(z 2) (b) x2(n) = x(2n) 8 -2k x(k)z X2(z)= x2(n)z n=-8 8 -n = n=-8 8 -n

x(2n)z x(k)z k 2 = k=-8 8 = x(k)+(-1)kx(k) k 2 ,k even z 2 k=-8 88 1 1 x(k)z k 2 1 2 )-k + x(k)(-z = 2 2 k=-8 k=-8

vv X( z + X(z) 1 = 2 3.14 (a) -1 13z X(z)= 1+3z-1 +2z-2 AB =+ (1+z-1) (1+2z-1) A =2,B = -1 Hence,x(n) = [2(-1)n (-2)n]u(n) (b) 1 X(z)= 1z-1 + 1 z-2 2A(11 z-1)+B(1 z-1) 22 = 1 z-2 1z-1 + 2 67 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

A =1,B =1 1 -1 1v (cos )z 4 2 Hence,X(z)= 12v1 (cos )z-1 +( v1)2z-2 41 22 -1 v (sin)z 4 2 +

12v1(cos )z-1 +( v1)2z-2 4 22 1 p 1 p )n Hence,x(n) =(v cos n +( v

)nsin nu(n) 24 24 (c) -6 -7 zz X(z)= + 1z-1 1z-1 x(n)= u(n 6)+u(n 7) (d) -2 1 z X(z)= +2 z-2 z-2 1+1+ 1 X(z) =2z-2 pp 1+ x(n)= cos nu(n)+2cos (n 2)u(n 2) 22 p x(n) =2d(n)cos nu(n) 2

(e) -1 -2 1 1+6z+ z X(z)= 4(12z-1 +2z-2)(11 z-1) 2 A(1z-1) Bz-1 C = ++ 12z-1 +2z-2 12z-1 +2z-2 11 z-1 2 3 23 17 A = ,B = ,C = 51020 31 p 23 1 p 17 1 Hence,x(n)= (v )n cos n + (v )nsin n + ( )n u(n) 5241024202 (f) -1

21.5z X(z)= 11.5z-1 +0.5z-2 11 = 11 z-1 + 1z-1 2 1 x(n) =( )n +1 u(n) 2 (g) -1 -2 1+2z+ z X(z)= 1+4z-1 +4z-2 . -2 2z-1 +3z =1(1+2z-1)(1+2z-1) 68 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 -2 2zz =1+ 1+2z-1 (1+2z-1)2 x(n)= d(n)2(-2)n-1 u(n 1)+(n 1)(-2)n-1 u(n 1) = d(n)+(n 3)(-2)n-1 u(n 1) (h) 11 1(z + )(z + ) 24 X(z)= 11 4(z 1)(z v j )(z v -j ) 2 2e4 2e 4 3 -11 -1 -2)z 1 (1+ z+ z 48 = 1 z-2)

4(11 z-1)(1z-1 + 22-1 z-1 Cz-1 A(11 z-1)zA(1 -1)z 22 = ++ 1z-1 + 11 z-2 1z-1 + z-2 11 z-1 222 173 A = ,B = ,C = 284 11 n-1 2 p 71 n-1 2 p 31 )n-1 Hence,x(n)= ( ) cos (n -

1)+ ( ) sin (n 1)+ ( u(n 1) 22482442 (i) -1 11 z 4 X(z)= 1 z-1 1+ 2-1 11 z = 1 z-11 z-1 1+ 41+ 22 1 11 )n-1 x(n) =()n u(n)+ (-

u(n 1) 242 (j) -1 1az X(z)= z-1 a 1 . 1az-1 . = a 11 z-1 a . -1 11 az = a 11 z-1 11 z-1 aa 11 1 )n-1 x(n)= ( )n u(n)+( u(n 1)

aa a 1)n+1 1)n-1 =(u(n)+( u(n 1) aa 3.15 -1 5z X(z)= (12z-1)(3z-1) 11 = 12z-1 + 11 z-1 3 1 If |z| > 2,x(n) =2n ( )n u(n) 3 69 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

If 1 3 < |z| < 2,x(n) If |z| < 1 3,x(n) = = -( 1 3)n u(-n 1)2n u(-n 1) ( 1 3)n u(-n 1)2n u(-n 1) 3.16 (a) x1(n) . X1(z) x2(n) = = = 1 4( 1 4)n-1 u(n 1) (1 4)z-1 11 4z-1,|z| > 1 4 . 1+ ( 1 2)n . u(n) . X2(z) Y(z) y(n) = = = = 1

1z-1 + 1 11 2z-1,|z| > 1 X1(z)X2(z) 4 3 11 4z-1 + 1 3 1z-1 + 1 11 2z-1 . 4 3( 1 4)n + 1 3 + ( 1 2)n . u(n) (b) x1(n) . X1(z) x2(n) . X2(z) Y(z) y(n) = = = = = = = u(n) 1 1z-1, d(n)+ ( 1 2)n u(n) 1+ 1

11 2z-1 X1(z)X2(z) 3 1z-1 1 11 2z-1 . 3( 1 2)n . u(n) (c) x1(n) . X1(z) x2(n) . X2(z) = = = = ( 1 2)n u(n) 1 11 2z-1, cospnu(n) 1+z-1 1+ 2z-1 + z-2 Y(z) = = X1(z)X2(z) 1+z-1 (11 2z-1)(1+ 2z-1 + z-2) 70 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(d) 3.17 3.18 (a) A(1+z-1) B = 1+2z-1 + z-2 + 11 z-1 2 21 A = ,B = 33 2 11 y(n)= cospn + ( )n u(n) 3 32 x1(n)= nu(n) -1 z . X1(z)= , (1z-1)2 x2(n) =2n u(n 1) -1 2z . X2(z)= 12z-1

Y(z)= X1(z)X2(z) -2 2z = (1z-1)2(12z-1) -1 -2 -2z2 = + 1z-1 (1z-1)2 12z-1 y(n)= -2(n +1)+2n+1. u(n) z + [x(n +1)] = z X+(z)x(0). = zX+(z)zx(0) 8 -n Therefore, zX+(z)= . x(n +1)z + zx(0) n=0 88 -n -n (z 1)X+(z) = . x(n)z + . x(n +1)z + zx(0) n=0 n=0 8 8 limz.1X+(z)(z 1) = x(0)+

. x(n +1) . x(n) n=0 n=0 = limm.8 [x(0)+x(1)+x(2)+... + x(m) -x(0)x(1)x(2)... x(m)] = limm.8x(m +1) = x(8) 8 8 -n . * )-n* x * (n)z = x(n)(z n=-8 n=-8 = X * (z * ) 71 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) 1[X(z)+X * (z * )] = 1 [z {x(n)} + z {x * (n)}] 22 = z x(n)+x *(n) 2 = z [Re{x(n)}] (c) x(n)x *(n) 2j = z [Im{x(n)}] 1 2j [X(z)X * (z * )] = z (d) 8 n -n Xk(z)= x( )z k n=-8,n/kinteger 8 . = m=-8 = X(z k) (e) -mk x(m)z 8 .

8 e jw0n -n x(n)z = -jw0z)-n x(n)(e n=-8 = n=-8 X(ze -jw0) 3.19 (a) X(z) Y(z) . y(n) Then,x(n) = = = = = = log(12z),|z| < 1 2 -z dX(z) dz -1 11 2z-1,|z| < 1 2 ( 1 2)n,n < 0 1 n y(n) 1 n ( 1 2)n u(-n 1) (b) X(z)

Y(z) = = log(11 2 z -1),|z| > 1 2 -z dX(z) dz 72 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 -1 z1 2 = ,|z| > 11z-12 21 1)n-1 Hence,y(n)= ( u(n 1) 221 x(n)= y(n) n 11 = ( )n u(n 1) n 2 3.20 (a) x1(n)= r nsinw0nu(n), 0 <r< 1 rsinw0z-1 X1(z)= 12rcosw0z-1 + r2z-2 jw0 Zero at z =0 and poles at z = re= r(cosw0

jsinw0). (b) z X2(z)= (1rejw0z-1)(1re-jw0z-1) z = 12rcosw0z-1 + r2z-2 (c)X1(z)and X2(z)differ by a constant, which can be determined by giving the val ue of X1(z) at z =1. 3.21 Assumethatthepolynomialhasreal coefficientsanda complexrootandprovethatthe compl ex conjugate of the root will also be a root. Hence, let p(z) be a polynomial and z 1 is a complex root. Then, nn-1 anz1 + an-1z + ... + a1z1 + a0 =0 (1) 1 The complex conjugateof(1) is ** * an(z1)n + an-1(z1)n-1 + ... + a1(z1)+a0 =0 * Therefore, z1 is also a root. 3.22 Convolution property: 8 8 z {x1(n)* x2(n)} = x1(k)x2(n k) z n=-8 k=-8 88 -n

= x1(k) x2(n k)z k=-8 n=-8 73 -n (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 = x1(k)z -kX2(z) k=-8 = X1(z)X2(z) Correlation property: r12(l)= x1(n)* x2(-n) R12(z)= z {x1(n)* x2(-n)} = X1(z)z {x2(-n)} -1) = X1(z)X2(z 3.23 23 zz X(z) =1+z + + + ... 2! 3! -2 -3 zz -1 +1+ z + + + ... 2! 3! 1 x(n)= d(n)+ n! 3.24 (a) 1 X(z)= 1+1.5z-1 0.5z-2

0.136 0.864 =+ 10.28z-1 1+1.78z-1 Hence, x(n) =[0.136(0.28)n +0.864(-1.78)n]u(n) (b) 1 X(z)= 10.5z-1 +0.6z-2 -1 -1 10.25z0.7326z = +0.3412 10.5z-1 +0.6z-2 10.5z-1 +0.6z-2 Then, x(n) =(0.7746)n [cos1.24n +0.3412sin1.24n]u(n) partial check: x(0) =1,x(1) =0.5016,x(2) = -0.3476,x(8) = 0. From difference equ ation, x(n)0.5x(n 1)+0.6x(n 2) = d(n)we obtain, x(0)=1,x(1)=0.5,x(2)= -0.35,x(8)=0. 3.25 (a) 1 X(z)= 11.5z-1 +0.5z-2 21 = 1z-1 10.5z-1 For |z| < 0.5,x(n) = [(0.5)n 2]u(-n 1) 74 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

For |z| > 1,x(n) =[2(0.5)n]u(n) For 0.5 < |z| < 1,x(n)= -(0.5)n u(n)2u(-n 1) (b) 1 X(z)= (10.5z-1)2 . -1 0.5z =2z (10.5z-1)2 For |z| > 0.5,x(n) =2(n +1)(0.5)n+1 u(n +1) =(n +1)(0.5)n u(n) For |z| < 0.5,x(n)= -2(n +1)(0.5)n+1 u(-n 2) = -(n +1)(0.5)n u(-n 1) 3.26 3 X(z)= 110 z-2 z-1 + 3 3 27 88 =+ 11z-1 13z-1 3 1 31 27

ROC: < |z| < 3,x(n)= ( )n u(n)3n u(-n 1) 3 838 3.27 8 -n X(z)= x(n)z n=-8 8 *-n = x1(n)x2(n)z n=-8 8 . 1 . -n = X1(v)v n-1dvx * 2(n)z 2pj c n=-8 . 8 . 1 . z * )-n -1 = X1(v)dv x2(n)( v 2pj v

c n=-8 . 8 #* . * 1 . z )-n = X1(v) x2(n)( v -1dv 2pj v * c n=-8 . * 1 X1(v)X2* ( z = )v -1dv 2pj v * c 3.28 Conjugation property: 75 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 . 8 #* . x * (n)z -n = . x(n)(z * )-n n=-8 n=-8 = X * (z * ) Parseval s relation: 8 . n=-8 x1(n)x * 2(n) = 8 . n=-8 1 2pj . c X1(v)v n-1dvx * 2(n) = 1 2pj . c X1(v) . 8 . x * 2(n)( 1 v )-n . v -1dv n=-8 1 X1(v)X2* ( 1 = )dv 2pj v * c 3.29 1 zndz x(n)= 2pj z a, c where the radius of the contour c is rc > |a|. For n< 0, let w = 1 . Then, z

. 1 -n-1 w 1 a x(n)= dw, 2pj w 1 ' c a ' 1 ' where the radius of c is 1 . Since < |a|, there are no poles within c and, hence x(n)=0for rc rc n< 0. 3.30 x(n)= x(N 1n), since x(n)is even. Then N-1 -n X(z)= x(n)z n=0 -1 -N+2 -N+1 = x(0)+x(1)z + ... + x(N 2)z + x(N 1)z N -1 2 hi -(N-1)/2(N-1-2n)/2 -(N-1-2n)/2 = zx(n) z+ zNeven

n=0 If we substitute z-1 for z and multiply both sides by z-(N-1) we obtain -(N-1)X(z -1) z = X(z) Hence, X(z)and X(z-1)have identical roots. This means that if z1 is root(or a ze ro) of X(z) * 1 then 1 is also a root. Since x(n)is real, then z1 must also be a root and so mus t * z1 z 1 3.31 From the definition of the Fibonacci sequence, y(n)= y(n 1) + y(n 2),y(0) =1. This is equivalent to a system described by the difference equation y(n)= y(n 1)+y(n 2)+x(n), 76 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

where x(n)= d(n) and y(n)=0,n< 0. The z-transform of this difference equation is Y(z)= z-1Y(z)+z-2Y(z)= X(z)Hence, for X(z)=1, we have 1 Y(z)= 1z-1 z-2 AB Y(z)= v + v 5+1 z-1 11- 5z-1 122 vv v 5+1 5115 where A = v ,B = v = -v 25 2525 vv vv 5+1 5+1 1515 Hence, y(n)= v ()n u(n)-v ()n u(n) 252 252 v . v 1 1+ 5)n+1 -

( 15)n+1 = v ( u(n) 52 2 3.32 (a) -1-2 Y(z)10.2z = X(z)10.3z -1 0.02z Y(z) (10.1z-1)(10.2z-1) = X(z)10.2z-1 -1 =10.1z (b) -1 Y(z)= X(z)10.1z Y(z) -1 =10.1z X(z) Therefore,(a) and(b) are equivalent systems. 3.33 1 X(z)= 1az-1 n . x1(n)= au(n) n

or x2(n)= -au(-n 1) Both x1(n) and x2(n) have the same autocorrelation sequence. Another sequence is obtained 1 from X(z-1)= 1-az -1) 1 X(z = 1az 1 =111z-1 a 1 Hence x3(n)= d(n)( )n u(n) a We observe that x3(n)has the same autocorrelation as x1(n)and x2(n) 77 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

3.34 -8 8 . -n . 2 -n )n H(z) =3n z +( z 5 n=-1 n=0 -11 2 =+ , ROC: < |z| < 3 13z-1 12 z-15 5 1 X(z)= 1z-1 Y(z)= H(z)X(z) 13 -1 z 5 = , ROC: 1 < |z| < 2 (1z-1)(13z-1)(12 z-1) 513 3 2

62 3 = -1z-1 13z-1 12 z-1 5 Therefore, 3 13 22 . y(n)= 3n u(-n 1)+ ( )n u(n) 2 635 3.35 (a) 1 h(n) =( )n u(n) 31 H(z)= 11 z-1 3 1 pn )n x(n) =( cos u(n) 23

-1 11 z 4 X(z)= 1 z-2 11 z-1 + 24 Y(z)= H(z)X(z) -1 11 z 4 = 1 z-2) (11 z-1)(11 z-1 + 324 v v 16 -1 3 -1 (11z33 z 7744 =+ + z-1 + 1z-2 z-1 + 1z-2

11z-1 11 711 32424 Therefore, . v . 11 61 pn 331 pn )n y(n)= ( )n + ( cos +( )nsin u(n) 73723723 (b) 1 h(n) =( )n u(n) 21 H(z)= 11z-1 2 1 1)-n x(n) =( )n u(n)+(

u(-n 1) 32 78 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) (d) (e) 11 X(z)= 11 z-1 12z-1 3 Y(z)= H(z)X(z) 5 -1 z 3 = (11z-1)(11z-1)(12z-1) 23 = 10 3 11 2z-1 + -2 11 3z-1 + -4 3 12z-1 Therefore, y(n) = 10 3 ( 1 2)n 2( 1 3)n . u(n)+ 4 3

2n u(-n 1) y(n)= -0.1y(n 1)+0.2y(n 2)+x(n)+x(n 1) -1 1+z H(z)= 1+0.1z-1 0.2z-2 1 x(n) =( )n u(n) 31 X(z)= 11z-1 3 Y(z)= H(z)X(z) -1 1+z = (11 z-1)(1+0.1z-1 0.2z-2) 328 -1 -8 33 =++ 11 z-1 10.4z-1 1+0.5z-1 3 Therefore, 1 28 2 11

)n y(n)= -8( )n + ( ( )n u(n) 33532 11 y(n)= x(n)x(n 1) 22 1 -1)X(z) . Y(z)= (1z 210 X(z)= z-2 1+ (1z-1)/2 Hence, Y(z) =10 z-2 1+ pn p(n 1) y(n) =5cos u(n)5cos u(n 1) 22

. pn pn . =5cos 5sin u(n 1)+5d(n) 22 10 pn p =5d(n)+ v sin( + )u(n 1) 2 2410 pn p = v sin( + )u(n) 2 24 y(n)= -y(n 2)+10x(n) 79 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Y(z) = 10 1+z-2X(z) X(z) = 10 1+z-2 Y(z) = 100 (1+z-2)2 = 50 1+jz-1 + 50 1jz-1 + -25jz-1 (1+jz-1)2 + 25jz-1 (1jz-1)2 Therefore, y(n) = {50[jn + (-j)n]25n[jn + (-j)n]} u(n) = (5025n)(jn + (-j)n)u(n) = (5025n)2cos pn 2 u(n) (f) h(n) = ( 2 5)n u(n) H(z) = 1 12 5z-1 x(n) = u(n)u(n 7) X(z) = 1z-n 1z-1 Y(z) = H(z)X(z) 1z-n = (12 5z-1)(1z-1) = 5 3 1-

z-1 + -2 3 12 5z-1 . 5 3 1z-1 + -2 3 12 5z-1 . z -7 Therefore, y(n) = 1 . 52( 2 )n . u(n)1 . 52( 2 )n-7 . u(n 7) 3 53 5(g) h(n) = ( 1 2)n u(n) H(z) = 1 11 2z-1 x(n) = (-1)n , 8 < n < 8 = cospn, 8 < n < 8 x(n)is periodic sequence and its z-transform does not exist. y(n) = |H(w0)|cos[pn+ (w0)],w0 1 = p H(z) = 11 2e-jw 1 H(p) =

1+ 1 2 2 = 3, = 0. 2 Hence, y(n) = 3cospn, 8 < n < 8 80 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(h) h(n) H(z) x(n) X(z) Y(z) Therefore, y(n) = = = = = = = = ( 1 2)n u(n) 1 11 2z-1 (n + 1)( 1 4)n u(n) 1 11 4z-1 + 1 4z-1 (11 4z-1)2 H(z)X(z) 1 (11 2z-1)(11 4z-1)2 4 11 2z-1 + -1 4 z-1 (11 4z-1)2 + -3 11 4z-1 . 4(

1 2)n n( 1 4)n 3( 1 4)n . u(n) 3.36 H(z) (a)Z1,2 (b)H(z) h(n) = = = = = 12z-1 + 2z-2 z-3 (1z-1)(11 2z-1)(11 5z-1) , 1 2 < |z| < 1 1z-1 + z-2 (11 2z-1)(11 5z-1) , 1 2 < |z| < 1 1 j v 3 2 , p1 = 1 2, p2 = 1 5 1+ . 5 2 11 2z-1

+ -2.8 11 5z-1 . z -1 d(n)+ . 5( 1 2)n 14( 1 5)n . u(n) 3.37 y(n) Y(z) = = 0.7y(n 1)0.12y(n 2)+x(n 1)+x(n 2) z-1 + z-2 10.7z-1 + 0.12z-2X(z) . x(n) X(z) Y(z) System is stable Y(z) = = = = nu(n) z-1 (1z-1)2 z-2 + z-3 (1z-1)2(13 10z-1)(12 50z-2) 4.76z-1 (1z-1)2 + -12.36 (1z-1) + -26.5 (13 10z-1) + 38.9 (1-

2 5z-1) 81 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

32 y(n) =4.76n 12.3626.5( )n +38.9( )n u(n) 105 3.38 Since the poles of H(z)areinsidethe unitcircle,thesystemisstable(polesat z = ). (a) 3 1 y(n) = 48y(n 1)y(n 2)+x(n) 1 Y(z) = 13 z-1 +481 z-2 X(z) Impulse Response: X(z) = 1 2 1 Y(z) = 2411z-1 11z-1 . h(n) = . . 2( 1 1 24)n ( )n u(n) 1 1 , 24 1 Step Response: X(z)= 1z-1 81

2 33 Y(z)= + 1z-1 11 z-1 11 z-1 24 8 1 11 . y(n)= 2( )n + ( )n u(n) 3 234 (b) 1 y(n)= y(n 1)y(n 2)+x(n)+x(n 1) 2 -1 1+z Y(z)= X(z) 1z-1 + 1 z-2 21j H(z)has zeros at z =0,1, and poles at z = . Hence, the system is stable. 2

Impulse Response: X(z) =1 v -13 -1 1( 2)-1cos z z 42 Y(z)= v + 1 z-2 12( 2)-1cos z-1 +( v1)2z-2 1z-1 + 2 4 2 1 . pp . . y(n)= h(n) =(v )n cos n + sin nu(n) 24 4 1 Step Response: X(z)= 1z-1 -1 1+z Y(z)= 1 z-2 (1-

z-1)(1z-1 + 21 -1 -(11 z-1) z2 22 = ++ 1 z-21 z-2 1z-1 1z-1 + 1z-1 + 22 1 . pp . y(n) =(v )n sin n cos nu(n)+2u(n) 24 4 (c) z-1(1+z-1) H(z)= (1z-1)3 82 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2 . h(n)= nu(n) Triple pole on the unit circle . the system is unstable. 1 Step Response: X(z)= 1z-1 z-1(1+z-1) Y(z)= (1z-1)4 -1 -2) -1) -1 1z-1(1+4z+ z1z-1(1+z1 z = + + 3 (1z-1)4 2 (1z-1)3 6(1z-1)2 111 3 2 y(n) =( n + n + n)u(n) 326 1 = n(n +1)(2n +1)u(n) 6 (d) y(n) =0.6y(n -

1)0.08y(n 2)+x(n) 1 Y(z)= X(z) 10.6z-1 +0.08z-2Impulse Response: X(z) =1 1 H(z)= (11z-1)(12z-1) 55 12 . zeros at z = 0, poles at p1 = ,p2 = system is stable. 25 -12 H(z)= + 11z-1 12z-1 55 21 )n ( . h(n) =2( )n u(n) 551 Step Response: X(z)= 1z-1 1 Y(z)= (11 z-1)(1-

2 z-1)(1z-1) 5525 1 4 12 4 3 Y(z)= ++ 1z-1 11 z-1 12 z-1 55 25 11 42 . )n y(n)= + ( ( )n u(n) 12 4535 (e) y(n) =0.7y(n 1)0.1y(n 2)+2x(n)x(n 2) -2 2z Y(z)= X(z) 10.7z-1 +0.1z-2-2 2z = X(z) (1-

1 z-1)(11 z-1) 52 11 zeros at z =0,2, and poles at z = , . Hence, the system is stable. 25 Impulse Response: X(z) =1 . -5 46 . 3 15 -1 H(z) =2+ + z 11 z-1 11 z-1 25 83 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

51 46 1 )n-1 )n-1 . h(n) =2d(n)( u(n 1)+ ( u(n 1) 321551 Step Response: X(z)= 1z-1 -2 2z Y(z)= (1z-1)(11 z-1)(11 z-1) 255 10 -23 23 6 =++ 1z-1 11 z-1 11 z-1 25 5 10 1 23 1 )n y(n)= + (

( )n u(n) 2 3265 3.39 (1+z-1)1 j X(z)= ,p= + (11 z-1)(1pz-1)(1p * z-1) 22 2 (a) x1(n)= x(-n +2) -1) X1(z)= z -2X(z z-2(1+z) = , ROC:|z| < 2 (11 z)(1pz)(1p * z) 2 Zero at z = -1, Poles at z =2, 11 , p * and z =0. p (b) jn 3 x2(n)= e x(n)

-3 j X2(z)= X(e z) j -1 1+e 3 z = j j j (11 e 3 z-1)(1pe 3 z-1)(1p * e 3 z-1) 2 All poles and zeros are rotated by in a counterclockwise direction. The ROC for X2(z)is 3 the same as the ROC of X(z). 3.40 x(n) = ( 1 2)n u(n)1 4( 1 2)n-1 u(n 1) 1 1 z-1 X(z) = 11 2z-1 411 2z-1 = 11 4z-1 11 2z-1 y(n) = ( 1 3)n u(n) 1

Y(z) = 11 3z-1 84 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) H(z)= Y(z)X(z) -1 11 z 2 = (11 z-1)(11 z-1) 43 3 2 = 11 4z-1 11 3z-1 h(n) = . 3( 1 4)n 2( 1 3)n . u(n) (b) -1 11 z 2 H(z)= 71 z-2 z-1 + 11212 71 1 y(n)= y(n -

1)y(n 2)+x(n)x(n 1) 12122 (c) Refer to fig 3.40-1. (d) The poles ofthe system are inside the unit circle. Hence, the system is stab le. x(n) 7/12 -1/12 -1/2 + + + z -1 z-1 y(n) Figure 3.40-1: 3.41 H(z) If a 2 1 4a2 p1,2 = < = 1 1+a1z-1 + a2z-2 0, there are two complex poles -a1 jp4a2 a2 1 2 85 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Refer to fig 3.41-1. 3.42 |p1,2|2 . a2 2 If a1 a4a2 p1,2 2 -a1 + a1 4a2 2 2 a -a1 1 4a2 2 . a1 a2 a1 + a2 H(z) p2!2 a1 4a2 a1 =( )2 + < 1 22 < 1 = 0, there are two real poles 2 -a1 a1 4a2 =

2 < 1 and > -1 < 1 and > 1 Figure 3.41-1: -11 -2 z+ z 2 = 2 z-2 13 z-1 + 525 86 real real real -1-2 1 2 complex a2 a 1 -1 a 1 +a2=-1 a 1-a 2=1 a2=1/4 a1 2 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) H(z) = z -1 . 7 2 11 5z-1 + 9 2 12 5z-1 . h(n) = . 7 2( 1 5)n-1 + 9 2( 2 5)n-1 . u(n 1) (b) Y(z)= H(z)X(z) 1 X(z)= 1z-1 25 7 -3 88 Y(z)= ++ 1z-1 11z-1 12z-1 55 25 71 2 . y(n)= + (

)n 3( )n u(n) 8855 (c)Determinetheresponse causedbytheinitial conditionsandaddittotheresponsein(b). 32 y(n)y(n 1)+ y(n 2) =0 525 -2 Y+(z)3 Y+(z)z -1 +1+ 2 Y+(z)z + z -1 +2=0 5 25 2 -1 11 z 2525 Y+(z)= (11z-1)(12z-1) 551 -12 25 25 =+ 11z-1 12z-1 55 . 1 1 12 2

)n y +(n)= ( ( )n u(n) 255255Therefore, the total step response is 25 33 1 87 2 . )n y(n)= + ( ( )n u(n) 8 2005255 3.43 -2 [aY (z)+X(z)]z = Y(z) -2 z Y(z)= X(z) 1az-2Assume that a> 0. Then 1 1 a + vvH(z)= a (1az-1)(1+ az-1) 111 11 v + v= -

+ a 2a 1az-1 2a 1+ az-1 11 v v h(n)= d(n)+ ( a)n +(a)n. u(n) a 2a 1 Step Response: X(z)= 1z-1 -2 z Y(z)= vv (1z-1)(1az-1)(1+ az-1) 87 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11 1 v v (a-1) 2(a- a) 2(a+ a) =+ v + v 1z-1 1az-1 1+ az-1 11 v 1 v y(n)= + v ( a)n + v (a)n u(n) a 1 2(a a)2(a + a) 3.44 y(n)= -a1y(n 1)+b0x(n)+b1x(n 1) b0 + b1z-1 Y(z)= X(z) 1+a1z-1 (a) b0 + b1z-1 H(z)= . h(n)= b0(-a1)n u(n)+b1(-a1)n-1 u(n 1) 1+a1z-1 -1 (b1 b0a1)z = b0 + . h(n)= b0d(n)+(b1 b0a1)(-a1)n-1 u(n 1)

1+a1z-1 (b) 1 Step Response: X(z)= 1z-1 b0 + b1z-1 Y(z)= (1z-1)(1+a1z-1) b0 + b1 1 a1b0 b1 1 =+ 1+a1 1z-1 1+a1 1+a1z-1 b0 + b1 a1b0 b1 y(n)= +(-a1)n u(n) 1+a1 1+a1 (c)Let us computethe zero-inputresponseandaddittotheresponsein(b). Hence, Y+(z)+a1 . z -1Y+(z)+A. =0 -a1A Y+(z)= 1+a1z-1 = -a1A(-a1)n u(n) . yzi(n) The total response to a unit step is b0 + b1 a1b0 b1 a1A(1+a1) y(n)=+ (-a1)n u(n) 1+a1 1+a1 (d) x(n)= cosw0nu(n) 1z-1cosw0 X(z)= 12z-1cosw0 + z-2

-1 (b0 + b1z-1)(1zcosw0) Y(z)= (1+a1z-1)(12z-1cosw0 + z-2) -1-1 AB(1zcosw0) C(zcosw0) =+ + 1+a1z-1 12z-1cosw0 + z-2 12z-1cosw0 + z-2 Then, y(n) =[A(-a1)n + Bcosw0 + Csinw0]u(n) 88 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

where A,B and C are determined from the equations A+ B = b0 (2cosw0)A+(a1 cosw0)B +(sinw0)C = b1 b0cosw0 A(a1 cosw0)B +(sinw0)C = -b1cosw0 3.45 y(n) Y(z) x(n) X(z) Y(z) Y(z) where A B Then y(n) The steady state response is limn.8y(n)= yss(n) 3.46 (a) H(z)= = H(z)|z=1 =1 . C = 1 = y(n 1)+4x(n)+3x(n 1) 2 -1 4+3z = X(z) 11z-1 2

jw0n = eu(n) 1 = 1ejw0z-1 -1 4+3z = (11z-1)(1ejw0z-1) 2 AB = 11 z-1 + 1ejw0z-1 2 5 = 1 ejw0 2 4ejw0 +3 = ejw0 1 2 = A( 1)n + Bejw0n u(n) 2 Bejw0n

= -j) (z rej)(z re C z(z +0.8) -12-2 12rcosz+ rz C (1+0.8z1) 1.8 =2.77 12rcos+r2 (b) The poles are inside the unit circle, so the system is stable. v (c)y(n)= -0.8y(n 1)+Cx(n)1.53Cx(n 1)+2.25Cx(n 2). Refer to fig 3.46-1. 3.47 (a) 2 -1 -2 X1(z)= z + z +1+ z + z 89 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) -0.8 -1.5 3 2.25 + ++ z -1 z -1 c y(n) x(n) -0.8 -1.5 3 2.25 + ++ z -1 z -1 c y(n) Figure 3.46-1: -1 -2 X2(z) =1+z + z Y(z)= X1(z)X2(z) -3 -4 = z 2 +2z +3+3z -1 +3z -2 +2z + z Hence, x1(n)* x2(n)= y(n) =1,2,3,3,3,2,1 . By one-sided transform: X+ -1 -2 (z) =1+z + z 1 X+ -1 -2 (z) =1+z + z 2 -3 -4 Y+(z) = 1+2z -1 +3z -2 +2z + z Hence, y(n)= {1,2,3,2,1} (b) Since bothx1(n)and x2(n)are causal,theone-sidedandtwo-sidedtransformyieldide ntical results. Thus,

Y(z)= X1(z)X2(z) 1 = (11 z-1)(11 z-1) 23 32 = 11z-1 11z-1 23 11 Therefore, y(n) =3( )n 2( )n u(n) 23 (c) By convolution, y(n)= x1(n)* x2(n) =4,11,20,30 ,20,11,4 . 90 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

By one-sided z-transform, X+ -2 (z) = 2+3z -1 +4z 1 X+ -1 (z) =2+z 2 Y+(z)= X+(z)X+(z) 12 -3 = 4+8z -1 +11z -2 +4z Therefore, y(n) =4,8,11,4 . (d) Bothx1(n)and x2(n)are causal. Hence, both types of transform yield the same result, i.e, -1 -2 -3 -4 X1(z) =1+z + z + z + z -1 -2 X2(z) =1+z + z Then, Y(z)= X1(z)X2(z) -5 -6 = 1+2z -1 +3z -2 +3z -3 +3z -4 +2z + z Therefore, y(n) =1,2,3,3,3,2,1 . 3.48 8 -n X+(z)= . x(n)z n=0 8 -n = z

n=0 1 = ,|z| > 1 1z-1 3.49 (a) . -1Y. -2Y-1 Y+(z)+ 1 z +(z)+y(-1)1 z +(z)+zy(-1)+y(-2)=0 2 4 (a) 1 -1 1 z 44 Hence, Y+(z)= 1 z-1 1 z-2 1+ 24 0.154 0.404 = 10.31z-1 1+0.81z-1 Therefore, y(n) =[0.154(0.31)n 0.404(0.81)n]u(n) (b) . -1Y. -2Y Y+(z)1.5z +(z)+1+0.5z +(z)+z -1 +0=0

-1 1.50.5z Y+(z)= 11.5z-1 +0.5z-2 20.5 = 1z-1 10.5z-1 Therefore, y(n) =[20.5(0.5)n]u(n) = 2(0.5)n+1. u(n) 91 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) . -1Y. 1 Y+(z)0.5z +(z)+1= 11z-1 3-1 1.516z Y+(z)= (11z-1)(10.5z-1) 37 2 2 = 10.5z-1 11z-1 3 7 1 Hence, y(n)= (0.5)n 2( )n u(n) 23 (d) . -2Y Y+(z)1 z +(z)+1= 1

41z-1 5 1 -1 z 44 Y+(z)= (1z-1)(11z-2) 44 -37 38 24 =++ z-11z-1 1z-1 1121+ 2 4 31 71 Hence, y(n)= ( )n + ()n u(n) 382242 3.50 If h(n)is real, even and has a finite duration 2N +1, then(with M =2N +1) -1 -2 -(M-1)/2 H(z)= h(0)+h(1)z + h(2)z + ... + h(M 1)z since h(n)= h(M n 1), then (M-1)/2 -(M-1)/2 H(z)= z -(M-1)/2(h(0) z+ z

(M-3)/2 -(M-3)/2 +h(1) z+ z + ... + h(M 1/2)) with M =2N +1, the expression becomes . N -N H(z)= z -N(h(0)z + z N-1 -(N-1) +h(1) z + z N-2 -(N-2) +h(2) z + z + ... + h(N)) N-1 N-1 -NN-n -(N-n) = zh(N)+ h(n)z + h(n)z n=0 n=0 -N-1). = zh(N)+P(z)+P(z Now, suppose z1 is a root of H(z), i.e., -N-1 H(z1)= zh(N)+P(z1)+P(z )=0 11 -1 Then, h(N)+P(z1)+P(z ) =0. 1 1 This implies that H( )=0 since we again have z1 -1 h(N)+P(z )+P(z1) =0. 1 92

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

3.51 (a) z 11 H(z)= , ROC: < |z| < 2 1 (z + )(z +3)(z 2) 2 2 (b) The system can be causal if the ROC is|z| > 3, but it cannot be stable. (c) ABC H(z)= ++ 1 z-1 1+3z-1 12z-1 1+ 2 (1) The system can be causal; (2) The system can be anti-causal; (3) There are t wo other noncausal responses.The corresponding ROC for each of these possibilities are : 1 ROC1 : |z| > 3; ROC2 : |z| < 3; ROC3 : < |z| < 2; ROC4 :2 < |z| < 3; 2 3.52 x(n)is causal. (a) 8

-n X(z)= x(n)z n=0 limz.8X(z)= x(0) (z-1)4 (b)(i)X(z)= 2. limz.8X(z)= 8. x(n)is not causal. (z-1)3 3 -2 (1-1z )2 (ii) X(z)= 2. limz.8X(z) = 1 Hence X(z) can be associated wih a causal 1-1z-1 3 sequence. (z-1)2 (iii) X(z)= 3. limz.8X(z) = 0. Hence X(z) can be associated wih a causal (z-1)3 2 sequence. 3.53 n1 Theansweris no. Forthegivensystem h1(n)= au(n). H1(z)= -1 ,|a| < 1. This system 1-az -33 naz is causal and stable. However when h2(n)= au(n +3) . H2(z)= the system is stable -1 1-az

but is not causal. 3.54 Initial value theorem for anticausal signals: If x(n)is anticausal, then x(0)= l imz.0X(z) P0 -n 2 Proof: X(z)= = x(0)+x(-1)z + x(-2)z+ ... Then limz.0X(z)= x(0) n=-8 x(n)z 3.55 1)n-2 s(n)=( u(n +2) 3 (a) h(n)= s(n)s(n 1) 93 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1)n-2 1)n-3 =( u(n +2) ( u(n +1) 33=34d(n +2) 54d(n +1) 18( 1 )n u(n) 3 -18 H(z) =81z 2 54z + z-1 11 381z(z-1) = 11z-1 3 1 H(z)has zeros at z =0,1 and a pole at z = 3. (b)h(n)=81d(n +2) 54d(n +1) 18(1)nu(n) 3 (c) The system is not causal, but it is stable sincethe pole is inside the unit circle. 3.56 (a) . n-1 1 z x(n)= 2pj 1-

1 z-1 dz c 2 . n 1 z = 2pj z 1 dz c 2 1 )n for n = 0,x(n) =( 2 11 for n< 0,x(-1) = 2pj c z(z 1) dz 2 11 = z 1 2 | |z=0 + z z= 1 2

= 0 11 x(-2) = 2pj c z2(z 1) dz 2 d 11 = dz z 1 2 | |z=0 + z2 z= 1 2 = 0 By continuing this process, we find that x(n)=0for n< 0. (b) 11 X(z)= ,|z| < 11z-12 2 . n 1 z1 x(n)=

dz, where c is contour of radius less than 2pj z 1 2 c 2 For n = 0, there are no poles enclosed in c and, hence, x(n)=0. For n< 0, we have 11 x(-1) = 2pj c z(z 1) dz 2 1 = |z=0 = -2 z 1 2 11 x(-2) = 2pj z2(z 1) dz c 2 d 1 = |z=0 = -4 dz z 1 2 94 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 Alternatively, we may change variables by letting w = z. Then, . -n 1 w1 x(n)= ()dw, 2pj c ' w-1 1 w2 2 1 -1 = wn+1(11 dw 2pj c ' w) 2 . -n-1 12w = dw 2pj w 2) ' c = -(2)-n , n< 0 (c) z a 1 X(z)= ,|z|

> 1az |a| 1 . n-1 z a 1 x(n)= z dz, c has a radius greater than 2pj 1az |a| c 1 . -1zn-1(z a) = 2pj a z 1 dz c a -11 1 )n-1( ( a) For n = 0,x(n)= aa a 1)n-1 ( 1)n+1 =( aa 1 . -1(z a) For n =0,x(n)= 2pj c az(z 1) dz

a -a 1 a -1 = -1 + a 1 a aa = -1(a 2 +1 a 2) a -1 = a For n< 0, we let w = z -1 . Then 1 . -w-n-1(w-1 a)1 x(n)= ()dw, 2pj 1aw-1 w2 ' c =0, for n< 0 (d) -1 11 z1

4 X(z)= ,|z| > 11z-1 1z-22 6637 10 10 1 z-1 = 11 z-1 + 1+ 23 . 3 n . 7 n z 1 z1 10 10 x(n)= dz 1 2pj c z 1 dz + 2pj c z + 23 1 where the radius of the contour c is greater than |z| = 2. Then, for n = 0 . 3 1 71 x(n)=

( )n + ()n u(n) 102103For n< 0,x(n) =0 95 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

3.57 2 1a1 X(z)= , a< |z| < ,0 <a< 1 (1az)(1az-1)a 1 -1 = 1az-1 + 11 z-1 a . n . n 1 z1 z x(n)= dz 2pj z adz 2pj z 1 cc a . n 1 z For n = 0, dz = a n and 2pj z a c . n

1 z dz =0 2pj z 1 c a . n 1 z For n< 0, dz = 0 and 2pj z a c . n 1 z -n dz = -a 2pj z 1 c a 3.58 20 z1 X(z)= , < |z| < 2 5 (z 1)(z 2)5(z +

)2(z +3) 2 22 . n-120 1 zz x(n)= 5 2pj c (z 1)(z 2)5(z + dz )2(z +3) 22 1 z x(-18) = 5 2pj (z 1)(z 2)5(z + dz )2(z +3) c 221 2 = 5 (1 2)5(1 + )2(1 +3) 2 222 1 2 = (3)5(3)2(7) 22-25

= (37)(7) -32 = 15309 96 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 4 4.1 (a)Since xa(t)is periodic, it can be represented by the fourier series 8 j2kt/ xa(t)= cke k=-8 . 1 j2kt/dt where ck = Asin(pt/t)e t 0 . A hi jt/ e -jt/-j2kt/dt = ee j2t 0 j(1-2k)t/ -j(1+2k)t/ Aee = j2tj(12k) --j(1+2k) 220 A 11 =

+ p 12k 1+2k 2A = p(14k2) . 8 k -j2(F)tdt Then, Xa(F)= xa(t)e -8 8 . 8 k )tdt -j2(F= ck e -8 k=-8 8 k = ckd(F ) t k=-8 Hence, the spectrum of xa(t)consistsof spectrallinesoffrequencies k,k =0,1,2,... w ith

amplitude |ck| and phases 6 ck. 1. 21. A2 (b)Px = x(t)dt = A2sin2(t )dt = 0a 0 2 (c) The power spectral densityspectrum is |ck|2,k =0,1,2,.... Refer to fig 4.1-1. (d) Parseval s relation . 1 2 Px = x (t)dt a t 0 = |ck|2 8 8 . 4A2 . 1 |ck|2 = p2 (4k2 1)2 k=-8k=-8 97 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim

itris G. Manolakis. ISBN 0-13-187374-1.

|c0| 2 . . . . . . . . .. . |c-2 |2 |c-1 |2 |c1|2 |c 2| 2 -1 0 1 2-2 k Figure 4.1-1: . 1+ 2 32 + 2 152 + . . . . Hence, 8 . k=-8 |ck|2 = = = = 4A2 p2 . 1+ 2 32 + 2 152 + . . . . 1.2337(Infinite series sum to 4A2 p2 (1.2337) A2 2 p2 8 ) 4.2 (a) xa(t) Xa(F) |Xa(F)| = = = = = Ae-at u(t), a > 0 . 8 0 Ae-at e -j2Ftdt

A -a j2pF e -(a+j2F)t 8 0 A a + j2pF A . a2 + (2pF)2 98 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1( 2pF Xa(F)= -tan) a Refer to fig 4.2-1 A = 2, a = 4 2 |Xa(F)| phase of Xa(F) 1 0 -1 0 -2 -10-5 0 510 -10-5 0 510 --> F --> F Figure 4.2-1: (b) . 8 . 8 Aeat -j2Ftdt+ Ae-at -j2Ftdt Xa(F)= ee 00 AA =+ a j2pF a + j2pF 2aA = a2 +(2pF)2 2aA |Xa(F)|

= a2 +(2pF)2 Xa(F) =0 Refer to fig 4.2-2 4.3 (a)Refer to fig 4.3-1. . |t| 1, |t|= t x(t)= 0, otherwise . 0 . t -j2Ftdt+ t -j2Ftdt Xa(F) = (1+ )e (1)e tt - 0 Alternatively, we may find the fourier transform of . 1 , -t<t = 0 y(t)= x ' (t)= 1 0 <t = t ,

99 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

A = 2, a = 6 0 5 10 15 20 25 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 ---> F |Xa(F)| Figure 4.2-2: x(t) |X(F)| t t -t 0 1/t 2/t F 0 Figure 4.3-1: 100 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Then, . -j2Ftdt Y(F)= y(t)e . 0 . 1 -1 -j2Ftdt+ -j2Ftdt = e ( )e tt - 0 2sin2pFt = jpFt 1 and X(F)= Y(F) j2pF sinpFt 2 = t pFt sinpFt 2 |X(F)| = t pFt Xa(F) =0 (b)

. Tp/2 1 -j2kt/Tpdt ck = x(t)e Tp -Tp/2 . 0 . 1 tt -j2kt/Tpdt+ -j2kt/Tpdt = (1+ )e (1)e Tp tt - 0 2 t sinpkt/Tp = Tp pkt/Tp 1 k (c)From(a)and(b),we have ck = Xa( ) Tp Tp 4.4 (a) x(n)= ..., 1,0,1,2,3,2,1,0,1,... . N =6 5

1 . -j2kn/6 ck = x(n)e 6 n=0 . -j2k -j2k -j4k -j10k . 633 6 = 3+2e + e + e +2e 1 pk 2pk = 3+4cos +2cos 6 33 94 1 4 Hence, c0 = ,c1 = ,c2 =0,c3 = ,c4 =0,c5 = 6666 (b) 5 1 . Pt = 2 |x(n)| 6 n=0 = 1(32 +22 +12 +02 +12 +22) 6 101

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

19 = 16 5 . 2 Pf = |c(n)| n=0 94 1 4 )2) =( )2 +( )2 +02 +( )2 +02 +( 1666619 = 16 Thus, Pt = Pf 19 = 16 4.5 1 x(n)=2+2cospn/4+cospn/2+ cos3pn/4,. N =8 2 (a) 7 1 . -jkn/4

ck = x(n)e 8 n=0 11 3vv vv . x(n)= 3 13 3 ,2+ 2,1,22, ,22,1,2+ 2 24 424 4 11 Hence, c0 =2,c1 = c7 =1,c2 = c6 = ,c3 = c5 = ,c4 =0 24 (b) 7 . 2 P = |c(i)| i=0 111 1 = 4+1+1+ +

+ + 4 4 16 16 53 = 8 4.6 (a) p(n 2) x(n) =4sin 3 2p(n 2) =4sin 6 5 1 . -2jkn/6 ck = x(n)e 6 n=0 5 4 . 2p(n 2) -2jkn/6 = sin e 66 n=0 1 hi

-j2k/3 e -jk/3 -jk/3 -j2k/3 = v-e + e + e 3 102 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 2pk pk -j2k/3 = v (-j2) sin + sin e 363 -j2/3 * Hence, c0 =0,c1 = -j2e ,c2 = c3 = c4 =0,c5 = c 1 and |c1| = |c5| =2,|c0| = |c2| = |c3| = |c4| =0 p 2p 5p c1 = p + = 23 6 -5p c5 = 6 c0 = 6 c2 = 6 c3 = 6 c4 =0 (b) 2pn 2pn x(n)= cos + sin . N =15 35 ck = c1k + c2k 2n where c1k is the DTFS coefficients of cos 35 and c2k is the DTFS coefficients of sin2n . But

2pn 1 j 3 2n -j2n (e + 3 ) cos = e 32 Hence, 1 2,k =5,10 c1k = 0, otherwise Similarly, 2pn 1 -j2n 5 j2n 5 sin = (e e ). 52j Hence, c2k = .. . 21 j,k =3 -1 ,k =12 2j 0, otherwise Therefore,

ck = c1k + c2k . ..... ..... 1 k =3 2j, 1 k =5 1 ,k =10 2, 2 -2 j 1 ,k =12 0, otherwise 2n 1 (c)x(n)= cos sin2n = sin16n 1sin4n Hence, N = 15. Following the same method 3 5215 215 . asin(b)above,we findthat .. . -1 ,k =2,7 4j 1 ,k =8,13 ck = 4j 0, otherwise (d)

N = 5 4 1 -j2nk 5 ck = 5 n=0 x(n)e 1 = e 5 -j2k -j4k -j6k -j8k 5 5 55 +2e 2e e 2j 2pk 4pk = -sin( )2sin( ) 5 5 5 103 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Therefore, c0 =0, 2j. 2p 4p . c1 = -sin( )+2sin( ) 55 5 2j. 4p 2p . c2 = sin( )2sin( ) 55 5 c3 = -c2 c4 = -c1 (e) N =6 5 1 . -j2nk ck = 6 x(n)e 6 n=0 1 . -jk -j2k -j4k -j5k . 333 3 = 1+2e e -

e +2e 6 1 pk 2pk = 1+4cos( )2cos( ) 6 33 1 Therefore, c0 = 2 2 c1 = 3 c2 =0 -5 c3 = 6 c4 =0 2 c5 = 3 (f) N =5 4 1 . -j2nk ck = 5 x(n)e 5 n=0 1 . -j2k .

5 = 1+e 5 2 pk -jk 5 = cos( )e 55 2 Therefore, c0 = 5 2 p -5 j c1 = cos( )e 5522p -5 j2 c2 = cos( )e 55 23p -5 j3 c3 = cos( )e 55 24p -5 j4 c4 = cos(

)e 55 (g)N =1 ck = x(0)=1 or c0 =1 (h) N =2 104 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 1 . -jnk ck = x(n)e 2 n=0 1 -jk) = (1e 2 . c0 =0,c1 =1 4.7 (a) 7 j2nk 8 x(n)= cke k=0 j2pk 8 Note that if ck = e, then 77 j2pk j2nk j2(p+n)k 88 8 ee = e k=0 n=0 =8,p= -n =0,p6-n =

1 . j 8 2k -j2k . 1 . j 8 6k -j6k . 8 Since ck = e + e 8 + e e 22j We have x(n) =4d(n +1)+4d(n 1)4jd(n +3)+4jd(n 3),-3 = n = 5 (b) vv vv 33 33 c0 =0,c1 = ,c2 = ,c3 =0,c4 = ,c5 = ,c6 = c7 =0 22 22 7 j2nk 8 x(n)= cke

k=0 v j2n 4 3 . jn = e 4 + e e j4n 4 e j5n 4 2 v . pn pn . j(3n-2) =3 sin 4 + sin e 24 (c) 4 j2nk x(n)= cke 8 k=-3 jn -jn 1 jn 1 -jn 1 j 4 3n 1 -j3n = 2+e 4 + e 4 + 2 + 4

e e 2 + e + e 22 44 pn pn 13pn = 2+2cos + cos + cos 4 224 4.8 (a) If k =0,N,2N,. . . N-1 N-1 j2kn/N e= 1= N n=0 n=0 105 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

If k 60,N,2N, = ... N-1 j2k . 1e j2kn/N e= 1ej2k/N n=0 =0 (b) Refer to fig 4.8-1. (c) k=1 k=2 k=3 s 2(4) s 1(2) s1(1) s 2(1) s (0) s (1) 3 s s1(3) 1(0) s2(0) 3s3(2) s (3) 3 s (3) s (4) (5) s1s2(2) 2s(5) 3 s1(4) 3 s (5) 2 k=4 k=5 k=6 s (5) 4

(5) s 5s (4) s 4(2) 5s (0) s (0) s(3) s5(0) 6. 5 4 . s (3) s (1) . s (1) 4 s (2) 5 4s6(5) 5 s (4) 4 Figure 4.8-1: N-1 N-1 * j2kn/N-j2in/N sk(n)si(n)= ee n=0 n=0 N-1 j2(k-i)n/N = e n=0 = N,k = i =0,k 6= i Therefore, the {sk(n)} are orthogonal. 4.9 (a) x(n)= u(n)u(n 6) 106

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 X(w)= x(n)e n=-8 5 -jwn = e n=0 -jwn = 1e-j6w 1e-jw (b) x(n) = 2n u(-n) 0 -jwn 2n e X(w)= n=-8 8 jw e )n ( = 2 m=0 2 = 2ejw (c) 1 x(n) =(

)n u(n +4) 4 8 1)n -jwn e X(w) =( 4 n=-4 8 1)m -jwm)44 j4w ee =( ( 4 m=0 44j4w e = 11 e-jw 4 (d) x(n)= ansinw0nu(n),|a| < 1 8 jw0n e-jw0n e -jwn an X(w)= e 2j n=0

8 8 in 1 in 1 ae-j(w-w0) ae-j(w+w0) = 2j 2j n=0 n=0 1 11 = 2j 1ae-j(w-w0) 1ae-j(w+w0) = asinw0e-jw 12acosw0e-jw + a2e-j2w (e) 8 x(n) = |a|nsinw0n,|a| < 1 8 n Note that |x(n)| = |a| |sinw0n| n=-8 n=-8 p Suppose that w0 = , so that |sinw0n| =1.

2 88 n=-8 n |a| = |x(n)|.8. n=-8 107 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Therefore, the fourier transform does not exist. (f) 2(1)n , |n|= 4 2 otherwise x(n)= 0, 4 -jwn X(w)= x(n)e n=-4 4 1 )n -jwn e =2( 2 n=-4 j4w 2e = 1e-jw 1 -j4w 3ej3w -j3w 2e-j2w ejw -jw j4w +4e + e j2w +2e + e -4e 2

j4w 2e =+ j[4sin4w +3sin3w +2sin2w + sinw] 1e-jw (g) 8 . X(w)= x(n)e n=-8 -jwn j2w ejw -j2w = -2e+ ejw +2e = -2j[2sin2w + sinw] (h) A(2M +1 -|n|), |n|= M 0, |n| >M x(n)= M X(w)= x(n)e n=-M M -jwn = A (2M +1 -|n|)e n=-M M -jwn -jwk jwk) + e (2M +1)A+ A (2M +1 k)(e =

k=1 M = (2M +1)A+2A (2M +1 k)coswk k=1 4.10 (a) . 1 jwndw x(n)= X(w)e 2p . -w0 . 11 jwndw jwndw = e+ e 2p 2p -w0 11 x(0) = (p w0)+ (pw0) 2p2p108 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

pw0 = p . -w0 1 jwndw jwn|-w0 For n 6=0,e= e jn 1 -jw0n e -jn) = (e jn . 1 jwndw jwn| e= e w0 jn w0 1 jn ejw0n) =

(e jnsinnw0 Hence, x(n)= ,n 60 = np (b) X(w)= cos 2(w) 11 -jw)2 jw =( e+ e 22 = 1(ej2w +2+ e -j2w) 4 . 1 jwndw x(n)= X(w)e 2p 1 = [2pd(n +2)+4pd(n)+2pd(n 2)] 8p 1 =

[d(n +2)+2d(n)+d(n 2)] 4 (c) . 1 jwndw x(n)= X(w)e 2p . w0+w 2 1 jwndw = e 2p w0-w 2 2 sin(ndw/2) jnw0 = dw e p ndw/2 (d) 1 (. /8 . 3/8 . 7/8 . jwndw+ jwndwjwndwjwndw x(n)= Re 2ee+ e+ e

2p 0 /86/87/8 1 ". /8 . 3/8 . 7/8 . . = 2coswndw + coswndw + coswndw +2coswndw p 0 /86/87/8 17pn 6pn 3pn pn = sin + sin sin sin np 8 8 88 4.11 x(n)+x(-n) xe(n)= 2 109 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

xo(n) Then, XR(w) jXI(w) Now, Y(w) y(n) 4.12 (a) 1 ". 9/10 1 1 = ,0,1,2,1,0, 2. 2 x(n)x(-n) = 2 11. = ,0,-2,0,2,0, 2. 2 3 -jwn = xe(n)e n=-3 3 -jwn = xo(n)e n=-3 j2w = XI(w)+XR(w)e. Therefore, j2w. = F-1 {XI(w)}

+ F-1XR(w)e= -jxo(n)+xe(n +2) 1 jj 1 j = ,0,1,2,1+ ,0, j2,0, 222. 22 . -8/10 . . -9/10 jwndwjwndw jwndw+2 jwndw+2 x(n)= e+ eee 2p 8/10 -9/10 9/10 1 . 1 j9n/10 e -j9n/10 ej8n/10 -j8n/10) = (e+ e 2p jn 2 j9n/10 -j9n/10 jn e -jn) + (-e+ e + e jn 1 = [sinpn sin8pn/10sin9pn/10] np

1 = [sin4pn/5+sin9pn/10] np (b) 1 x(n)= 2p 1 = 2p 1 = 2p 1 = pn (c) . 0 . 1 jwndw+ jwndw X(w)e X(w)e 2p - 0 . 0 . w 1 w jwndw jwndw

( +1)e+ e p 2pp - 0 . wejwn . jwn| |0 e- + -p jnp jn pn -jn/2 sin e 2 . wc+ww . -wc+ 1 1 jwndw + 2 2 2e jwndw x(n)= 2e ww 2p 2p -wc-

wc 2 2 . jwn . w e-wc+w 11 wc+ 2 2 + | jn = e jwn| wc p jnp w 2 w 2 -wc110 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

)n ww ww j(wc+ )n ej(wc)n -j(wc+ e )n e-j(wc+ 2 2 2 2 2 e = = 4.13 X2(w)= = = X(w)= = = = = 4.14 (a)X(0)= (b)6 X(w)= 1 (c)x(0)= 2 (d) x(n)=

n p for all . pn 2j 2 . ww . sin(wc + )n sin(wc )n pn 22 1, 0 = n = M x1(n)= 0, otherwise M -jwn X1(w)= e n=0 -jw(M+1) 1e = 1e-jw 1, -M = n =-1 x2(n)= 0, otherwise -1 -jwn e

n=-M M jwn e n=1 jwM 1e jw e 1ejw X1(w)+X2(w) jw ejw 1e-jw(M+1) ejw(M+1) jwM -jwM 1+e+ e+ e X(w)dw Hence, X(w)dw =2px(0)= -6p - 8 -jn X(p)= . x(n)e = . (-1)n x(n)= -342 n=-8 n 2e-jw ejw 2coswM 2cosw(M +1) 22cosw w 2sin(wM + w )cos 22 w

2sin22 1 sin(M + )w 2 sin(w ) 2 -1 w . = -9 . 2pP (e)|X(w)|2dw = |x(n)|2 =(2p)(19)=38p -n 111 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

4.15 (a) (b) See fig 4.15-1 X(w) = . n x(n)e -jwn X(0) = . n x(n) dX(w) dw |w=0 = -j . n nx(n)e -jwn|w=0 = -j . n nx(n) Therefore, c = jdX(w) dw |w=0 X(0) X(0)= 1 Therefore, c = 0 1 = 0. dX(w) dw w Figure 4.15-1: 4.16 n x1(n) = au(n) F1 . 1ae-jw 112 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Now, suppose that (n + k1)! n xk(n)= au(n) n!(k1)! F1 . (1ae-jw)k holds. Then (n + k)! n xk+1(n)= au(n) n!k! n + k = xk(n) k 1 XX -jwn -jwn Xk+1(w)= nxk(n)e + xk(n)e k nn 1 dXk(w) = j + Xk(w) kdw -jw ae1 =+ (1-

ae-jw)k+1 (1ae-jw)k 4.17 (a) -jwn -j(-w)n) * x * (n)e =( x(n)e = X * (-w) nn (b) 8 -jwn jwn x * (-n)e = x * (n)e= X * (w) nn=-8 (c) . -jwn . -jwn . -jwn y(n)e = x(n)ex(n 1)e n nn -jw Y(w)= X(w)+X(w)e -jw)X(w) = (1e (d) n y(n)= x(k) k=-8 = y(n)y(n 1) = x(n) -jw)Y(w) Hence, X(w) =(1e X(w) . Y(w)=

1e-jw (e) -jwn Y(w)= x(2n)e n -j n 2 = x(n)e w n w = X( ) 2 113 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(f) . n -jwn Y(w)= x( )e 2 n -j2wn = x(n)e n = X(2w) 4.18 (a) -jwn X1(w)= x(n)e n j2w -jw -j2w = e+ ejw +1+ e + e = 1+2cosw +2cos2w (b) -jwn X2(w)= x2(n)e n j4w -j2w -j4w = e+ ej2w +1+ e + e = 1+2cos2w +2cos4w (c) -jwn X3(w)= x3(n)e n j6w -j3w -j6w = e+ ej3w +1+ e + e

= 1+2cos3w +2cos6w (d)X2(w)= X1(2w)and X3(w)= X1(3w). Refer to fig 4.18-1 (e) If . x(nn ), an integer kk xk(n)= 0, otherwise Then, -jwn Xk(w)= xk(n)e n n, k an integer -jkwn = x(n)e n = X(kw) 4.19 (a) 1 jn/4 -jn/4)x(n) x1(n)= (e+ e 21 . pp . X1(w)= X(w )+X(w + ) 244

114 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

X(w) X(w) 1 2 p -0.5p 0.5p p w -p 0 0 X(w) 3 w -p/3 0 p/3 p Figure 4.18-1: (b) x2(n) = 1 2j(ejn/2 + e -jn/2)x(n) 1 . p p . X2(w) = 2j X(w 2)+X(w + 2) (c) x3(n) = 1 2(ejn/2 + e -jn/2)x(n) 1 . p p . X3(w) = 2 X(w 2)+X(w + 2) (d) x4(n) = 1 2(ejn + e -jn)x(n) 1 X4(w) = 2 [X(w p)+X(w + p)] = X(w p) 4.20 cy k = 1

N N-1 . y(n)e -j2kn/N n=0 115 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N-1. 8 . 1 X. -j2kn/N = x(n lN) e N n=0 l=-8 8 N-1-lN 1 X. -j2k(m+lN)/N = x(m)e N l=-8 m=-lN 8 N-1-lN -jw(m+lN) But x(m)e = X(w) l=-8 m=-lN 12pk Therefore, cy = X( ) k NN 4.21 sinwcn Let xN(n)= , N = n =

N pn = x(n)w(n) sinwcn where x(n)= , -8= n =8 pn w(n) =1, N = n = N =0, otherwise sinwcn F Then . X(w) pn =1, |w|= wc =0, otherwise XN(w)= X(w)* W(w) . = X()W(w )d . wc sin(2N +1)(w )/2 = d sin(w )/2 -wc 4.22 (a) -jwn X1(w)= x(2n +1)e n -jwk/2 jw/2 = x(k)ee k w

jw/2 = X( )e 2 jw/2 e = 1aejw/2 (b) n/2 -jwn X2(w)= x(n +2)ee n -jk(w+j/2)j2w = x(k)ee k 116 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

jp j2w = -X(w + )e 2 (c) -jwn X3(w)= x(-2n)e n -jkw/2) = x(k)e k w = X() 2 (d) . 1 j0.3n -j0.3n)x(n)e -jwn X4(w)= (e+ e 2 n 1 X. i -j(w-0.3)n -j(w+0.3)n = x(n) e + e 2 n 1

= [X(w 0.3p)+X(w +0.3p)] 2 -jw. -jw (e)X5(w)= X(w)X(w)e= X2(w)e (f) X6(w)= X(w)X(-w) 1 = (1ae-jw)(1aejw) 1 = (12acosw + a2) 4.23 . -jwn P (a)Y1(w)= y1(n)e= x(n)e-jwn The fourier transform Y1(w)can easily be n n,n even obtainedby combiningtheresultsof(b)and(c). (b) y2(n)= x(2n) -jwn Y2(w)= y2(n)e n -jwn = x(2n)e n -jwm/2 = x(m)e m w

= X( ) 2 Refer to fig 4.23-1. (c) . x(n/2),n even y3(n)= 0, otherwise -jwn Y3(w)= y3(n)e n 117 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Y(w) 2 p/2 2p -p -p/2 0 p 3p/2 Figure 4.23-1: -jwn = x(n/2)e n even -j2wm = x(m)e m = X(2w) We now return to part(a). Note that y1(n)may be expressed as . y2(n/2),n even y1(n)= 0,n odd Hence, Y1(w)= Y2(2w). Refer to fig 4.23-2. Y(w) 3 -p/8 p/8 p/2 7p/8 p -p -7p/8 0 Y(w) 1 -p -3p/4 -p/2 -p/4 0 3p/4 p/4 p/2 p Figure 4.23-2: 118 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 5 5.1 (a) Because the range of n is (-8,8), the fourier transforms of x(n) and y(n) do not exist. However, the relationship implied by the forms of x(n) and y(n) is y(n)= x3(n). In this case, the system H1 is non-linear. (b) In this case, 1 X(w) = 11 2e-jw , 1 Y(w) = 11 8e-jw , Hence, H(w) = Y(w) X(w) = 11 2e-jw 11 8e-jw . System is LTI Note however that the system may also be nonlinear, e.g., y(n)= x3(n). (c) and (d). Clearly, there is an LTI system that produces y(n) when excited by x(n), e.g. H(w)=3,forall w, or H()=3. 5 (e) Ifthis system is LTI, the period of the output signal would be the same as t he period of the input signal, i.e., N1 = N2. Since this is not the case, the system is nonlinear . 5.2 (a) WR(w) = M . n=0 wR(n)e -jwn = M .

e -jwn n=0 = = 1e-j(M+1)w 1e-jw e -jMw/2sin(M+1 2 )w sinw 2 119 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) LetwT(n)= hR(n)* hR(n 1), hR(n)= 1, 0 = n = M 0, otherwise 2 1 Hence, WT(w)= H -jw R(w)e !2 sinM sinw 2 4 2 w -jwM/2 = e (c) Let c(n) = 1 2(1+cos 2pn M ) Then, C(w) = p d(w)+ 1 2 d(w 2p M)+ 1 2 d(w + 2p

M) -p =w =p Wc(w) = 1 2p Z c()WR(w -)d = 1 2 WR(w)+ 1 2 WR(w 2p M)+ 1 2 WR(w + 2p M) Refer to fig 5.2-1 -2p/M+1 0 2p/M+1 -4p/M 0 4p/M -2p/M -2p/M+1 0 2p/M+1 2p/M |W(w)| R |W(w)| T |W(w)| c w w w Figure 5.2-1: 5.3 (a) 1 h(n) =( )n u(n) 2 120 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

H(w)= = = |H(w)| = = H(w)= = (b)(1) For the input x(n)= = X(w)= Y(w)= = y(n)= (2) x(n) First, determine xe(n) and xo(n) Then, XR(w) XI(w) |H(w)| (w) and Y(w) 8 . 1)n -jwn ( e 2 n=0 8 . 1 -jw)n ( e 2

n=0 1 11 e-jw 2 1 2 (11 cosw)2 +(1sinw)21 22 1 1 5 cosw 2 4 1 sinw -12 -tan 11 cosw 2 (w) 3p cos n 10 1 j3n 10 -j3n 10

(e + e ) 2 3p 3p pd(w )+d(w + ) ,|w|= p 1010 H(w)X(w) 3p 3p 3p H( )pd(w )+d(w + ) 1010103p 3pn 3p . |H( )|cos + ( ) 1010 10 = ..., 1,0,0,1,1,1,0,1,1,1,0,... . x(n)+x(-n) = 2 x(n)x(-n) = 2 -jwn

= xe(n)e n -jwn = xo(n)e n X2 . = R(w)+XI 2(w), XI(w) -1 = tan XR(w) = H(w)X(w) 121 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.4 (a) x(n)+x(n 1) y(n)= 2 1 -jw)X(w) Y(w)= (1+e 21 -jw) H(w)= (1+e 2 w -jw/2 =(cos )e 2 Refer to fig 5.4-1. (b) 1 0.8 -4-3-2-1 0 1 2 3 4 --> |H(w)| 0.6 0.4 0.2 0 --> w 2 -4-3-2-1 0 1 2 3 4 --> theta(w)

1 0 -1 -2 --> w Figure 5.4-1: x(n)x(n 1) y(n)= 2 1 -jw)X(w) Y(w)= (1e 21 -jw) H(w)= (1e 2 w -jw/2 j/2 =(sin )ee 2 Refer to fig 5.4-2. (c) y(n) = x(n + 1) x(n 1) 2 122 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.5 1 1.5 --> w 2 2.5 3 3.5 0 Refer to fig 5.4-3. (d) Refer to fig 5.4-4 (e) Refer to fig 5.4-5. (f) 0.5 1 Y(w) H(w) y(n) Y(w) H(w) y(n) Y(w) H(w) 1.5 2 2.5 --> w Figure 5.4-2: = 1 2(ejw e -jw)X(w) = 1 2(ejw e -jw) = (sinw)ej/2 = x(n + 1)+ x(n 1) 2 = 1 2(ejw + e -jw)X(w) = 1 2(ejw + e -jw) = cosw = x(n)+x(n 2)

2 = 1 2(1+e -j2w)X(w) = 1 2(1+e -j2w) = (cosw)e -jw 123 3 3.5 0 0.5 1 1.5 2 --> theta(w) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 3 2.5 2 1.5 1 0.5 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-3: 0 0.5 1 1.5 2 2.5 3 3.5 --> w 4 --> theta(w) 0 0.2 0.4 0.6 0.8 1 --> |H(w)| --> theta(w) 3 2 1 0 --> w Figure 5.4-4: 124 0 0.5 1 1.5 2 2.5 3 3.5 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.2 0.4 0.6 0.8 1 --> |H(w)| 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-5: x(n)x(n 2) y(n)= 2 Y(w)= 1(1e -j2w)X(w) 21 -j2w) H(w)= (1e 2 -jw+j/2 =(sinw)e 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 --> theta(w) Refer to fig 5.4-6 (g)

x(n)+x(n 1)+x(n 2) y(n)= 3 1 -jw -j2w)X(w) Y(w)= (1+e + e 31 -j2w) -jw H(w)= (1+e + e 31 jw -jw)e -jw = (1+e+ e 31 -jw = (1+2cosw)e 31 |H(w)| = | (1+2cosw)| 3 -w, 1+2cosw > 0 6 H(w)= p w, 1+2cosw < 0 Refer to fig 5.4-7. (h) 125 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-6: 0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 1 0 0.5 1 1.5 2 2.5 3 3.5 0 -1 -2 -3 --> w Figure 5.4-7: 126 --> theta(w) --> theta(w) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

y(n)= x(n)x(n 8) Y(w) =(1e -j8w)X(w) -j8w) H(w) =(1e j(/2-4w) = 2(sin4w)e Refer to fig 5.4-8. (i) 2 1.5 0 0.5 1 1.5 2 2.5 3 3.5 --> |H(w)| 1 0.5 0 --> w 2 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-8: y(n) =2x(n 1)x(n 2) -jw e -j2w)X(w) Y(w) =(2e -jw e -j2w)

H(w) =(2e =2cosw cos2w j(2sinw sin2w) . 2 |H(w)| = (2cosw cos2w)2 +(2sinw sin2w)21 2sinw sin2w . -1 (w) = -tan 2cosw cos2w --> theta(w) Refer to fig 5.4-9. (j) y(n) Y(w) H(w) = = = = x(n)+x(n 1)+x(n 2)+x(n 3) 4 1 4(1+e -jw + e -j2w + e -j3w)X(w) 1 3 . e -jw(ejw + e -jw)+e -j2w(ejw + e -jw). 1 2(e -jw + e -j2w)cosw 127 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 3 2.5 2 1.5 1 --> w 2 0 0.5 1 1.5 2 2.5 3 3.5 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-9: w -j3w/2 =(cosw)(cos )e 2 --> theta(w) Refer to fig 5.4-10. (k) y(n) Y(w) H(w) = = = = x(n)+ 3x(n 1)+ 3x(n 2)+x(n 3) 8 1 8(1+ 3e -jw + 3e -j2w + e -j3w)X(w) 1 8(1+e -jw)3 (cosw/2)3 e -j3w/2 Refer to fig 5.4-11. (l) y(n)

Y(w) H(w) |H(w)| (w) = = = = = x(n 4) e -j4wX(w) e -j4w 1 -4w Refer to fig 5.4-12. (m) y(n) Y(w) H(w) = = = x(n + 4) ej4wX(w) ej4w 128 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.2 0.4 0.6 0.8 1 --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 --> |H(w)| --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-10: 1 0.8 0.6 0.4 0.2 0 --> w 2 0 0.5 1 1.5 2 2.5 3 3.5 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-11:

129 --> theta(w) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-12: |H(w)| =1 (w) =4w Refer to fig 5.4-13. (n) y(n) Y(w) H(w) = = = = x(n)2x(n 1)+x(n 2) 4 1 4(12e -jw + e -j2w)X(w) 1 4(1e -jw)2 (sin2w/2)e -j(w-) Refer to fig 5.4-14. 5.5 (a) y(n) Y(w) H(w) = =

= x(n)+x(n 10) (1+e -j10w)X(w) (2cos5w)e -j5w Refer to fig 5.5-1. (b) 130 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 --> |H(w) | --> |H(w)| --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.4-13: 1 0 0.5 1 1.5 2 2.5 3 3.5 0.8 0.6 0.4 0.2 0 --> w 4 0 0.5 1 1.5 2 2.5 3 3.5 3 2 1 0 --> w Figure 5.4-14: 131 --> theta(w) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 2 1.5 1 0.5 0 --> w 2 0 0.5 1 1.5 2 2.5 3 3.5 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.5-1: p H( ) =0 10 --> theta(w) p 5p 5 -j 3 H( ) =(2cos )e 33 5p pp 5p y(n) =(6cos )sin( +

) 3 310 3 5pp 47p = (6cos )sin( ) 3 330 (c) H(0) =2 4p H( ) =2 10 2pn p y(n) = 20+10cos + 5.6 h(n) H(w) p Steady State Response: H( ) 2Therefore, yss(n) 52 = d(n)+2d(n 2)+d(n 4) -j2w -j4w = 1+2e + e -j2w)2

= (1+e = 4(cosw)2 e -j2w =0 =0,(n = 4) 132 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Transient Response: n (n-2) (n-4) 2 2 ytr(n) =10e 2 u(n)+20eu(n 2)+10eu(n 4) = 10d(n)+j10d(n 1)+10d(n 2)+j10d(n 3) 5.7 (a) y(n)= x(n)+x(n 4) Y(w) = (1+e -j4w)X(w) -j2w H(w) =(2cos2w)e Refer to fig 5.7-1. (b) 2 1.5 0 0.5 1 1.5 2 2.5 3 3.5 --> w --> |H(w)| 1 0.5 0 2 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.7-1: --> theta(w) ppp p

y(n)= cos n + cos n + cos (n 4)+cos (n 4) 2424 p pp But cos 2(n 4) = cos 2 ncos2p+ sin 2 nsin2p p = cos n 2 p p p and cos 4(n 4) = cos 4 ncosp sin 4 nsinp p = -cos n 4 p Therefore, y(n) = 2cos n 2 (c) Note thatH()=2 and H()=0. Therefore,the filterdoes notpassthesignal cos(n). 24 4 133 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.8 1 y(n) = 2 [x(n)x(n 2)] Y(w) = 1 2(1e -j2w)X(w) H(w) = 1 2(1e -j2w) =(sinw)ep j( 2 -w) H(0) Hence, yss(n) ytr(n) 5.9 x(n)= Acos n 4

(a)y(n)= x(2n)= Acos n . w = 22 2p 11 A2 (b)y(n)= x2(n)= A2cosn = 42A2 + 2 (c) y(n)= = = Hence, w =

5.10 (a) y(n)= Y(w)= H(w)= =0,H( )=1 2 p =3cos( n +60o) 2 =0 p cos 2n. Hence, w =0 and w = 2 x(n)cospn p Acos ncospn 4 A 5pA 3p cos n + cos n 24 24 3p 5p and w = 44 1

[x(n)+x(n 1)] 2 1 -jw)X(w) (1+e 21 -jw) (1+e 2 w -j = cos( )e 2Refer to fig 5.10-1. (b) 1 w 2 y(n) = [x(n)x(n 1)] 2 1 -jw)X(w) Y(w) = (1e 2 w |H(w)| = sin 2 (w)= e j(

2 w 2 ) 134 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 1 0.8 0.6 0.4 0.2 0 0 0.5 1 1.5 2 2.5 3 3.5 --> w 0 0.5 1 1.5 2 2.5 3 3.5 -2 -1.5 -1 -0.5 0 --> theta(w) --> w Figure 5.10-1: Refer to fig 5.10-2. (c) y(n) Y(w) H(w) = = = = 1 8 [x(n)+ 3x(n 1)+ 3x(n 2)+x(n 3)] 1 8(1+e -jw)3X(w) 1 8(1+e -jw)3 cos 3( w 2)e -j3w 2 Refer to fig 5.10-3. 135 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.2 0.4 0.6 0.8 1 --> |H(w)| --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.10-2: 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 0 0.5 1 1.5 2 --> theta(w) 0 0.2 0.4 0.6 0.8 1 --> |H(w)| --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.10-3: 136 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.11 y(n)= x(n)+x(n M) -jwM)X(w) Y(w) = (1+e -jwM) H(w) = (1+e wM 1 H(w) =0, at =(k+ )p, k =0,1,... 22or w = (2k+1)p/M, k =0,1,... wM |H(w)| = |2cos | 2 5.12 y(n)=0.9y(n 1)+bx(n) (a) Y(w) =0.9e -jwY(w)+bX(w) Y(w) H(w)= X(w) b = 10.9e-jw |H(0)| =1,. b = 0.1 . wM wM , cos > 0 22 (w)=

p wM wM , cos < 0 22 (b)|H(w0)|2 = 1 b2 1 . w0 =0.105 . = 21.81-1.8cosw0 2 (c) The filter is lowpass. 1 (d) For|H(w0)|2 = . w0 =3.036. This filter is a highpass filter. 2 5.13 (a) N-1 1 . |x(n)|2 Px = N n=0 N-1 = |ck|2 k=0 N -1 2 2 = c0 +2 |ck|2 k=1 Spurious power = Px 2|ck0|2 Px 2|ck0|2 THD =

Px 2|ck0|2 =1Px 137 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 (b) forf0= , refer to fig 5.13-1 96 1 for f0= , refer to fig 5.13-2 32 1 for f0= , refer to fig 5.13-3 256 1 (c) forf0= , refer to fig 5.13-4 96 1 for f0= , refer to fig 5.13-5 32 1 for f0= , refer to fig 5.13-6 256 The total harmonic distortion(THD) reduces as the number of terms in the Taylor approximation is increased. terms= 2 terms= 3 terms= 4 -20 -10 0 10 --> x(n) -50 0

50 --> x(n) -40 -20 0 20 --> x(n) 0 50 1000 50 1000 50 100 terms= 5 terms= 6 terms= 7 -10 0 10 20 --> x(n) -5 0 5 --> x(n) -2 0 2 4 --> x(n) 0 50 1000 50 1000 50 100 terms= 8 1 --> x(n) 0 -1 -2 Figure 5.13-1: 0 50 100 138 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

terms= 2 terms= 3 terms= 4 -20 -10 0 10 --> x(n) -20 0 20 40 --> x(n) -40 -20 0 20 --> x(n) 0 20 400 20 400 20 40 terms= 5 terms= 6 terms= 7 -10 0 10 20 --> x(n) -4 -2 0 2 --> x(n) -1 0 1 2 --> x(n) 0 20 400 20 400 20 40 terms= 8 -2 -1 0 1 --> x(n) 0 20 40 Figure 5.13-2: terms= 2 terms= 3 terms= 4 10 50 20 0

-10 --> x(n) 0 --> x(n) 0 -20 -20 -50 -40 0 100 200 300 0 100 200 300 0 100 200 300 terms= 5 terms= 6 terms= 7 40 5 4 20 0 --> x(n) 0 -5 --> x(n) 2 0 -20 -10 -2 0 100 200 300 0 100 200 300 0 100 200 300 terms= 8 1

0 -1 -2 0 100 200 300 Figure 5.13-3: 139 --> x(n) --> x(n) --> x(n) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

orig cos thd=2.22e-16 terms=2 thd=0.06186 terms=3 thd=0.4379 terms=2 thd=0.06186 terms=3 thd=0.4379 -200 -100 0 100 psd -400 -200 0 200 psd -100 -50 0 50 psd 0 0.5 10 0.5 10 0.5 1 terms=4 thd=0.5283 terms=5 thd=0.6054 terms=6 thd=0.6295 -100 -50 0 50 psd -100 -50 0 50 psd -100 -50 0 50 psd 0 0.5 10 0.5 10 0.5 1 terms=7 thd=0.06924 terms=8 thd=0.002657 -200 -100 0 100 psd -200 -100 0 100 psd 0 0.5 1 0 0.5 1 Figure 5.13-4: orig cos thd=0 terms=2 thd=0.07905 terms=3 thd=0.4439 -100 -50 0 50 psd -100

-50 0 50 psd -100 -50 0 50psd 0 0.5 10 0.5 10 0.5 1 terms=4 thd=0.5312 terms=5 thd=0.5953 terms=6 thd=0.6509 -100 -50 0 50 psd -100 -50 0 50 psd -100 -50 0 50 psd 0 0.5 10 0.5 10 0.5 1 terms=7 thd=0.05309 terms=8 thd=0.001794 -100 -50 0 50 psd -100 -50 0 50 psd 0 0.5 10 0.5 1 Figure 5.13-5: 140 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

orig cos thd=-6.661e-16 terms=2 thd=0.05647 terms=3 thd=0.4357 -200 -100 0 100 psd -200 -100 0 100 psd -200 -100 0 100 psd 0 0.5 10 0.5 10 0.5 1 terms=4 thd=0.5271 terms=5 thd=0.6077 terms=6 thd=0.6238 -200 -100 0 100 psd -200 -100 0 100 psd -200 -100 0 100 psd 0 0.5 10 0.5 10 0.5 1 terms=7 thd=0.07458 terms=8 thd=0.002976 -200 -100 0 100 psd -200 -100 0 100 psd 0 0.5 10 0.5 1 Figure 5.13-6: 5.14 (a)Refer to fig 5.14-1 1 (b)f0 =

50 --> xq(n) -0.5 -1 -1 -1.5 0 100 200 300 0 100 200 300 --> n --> n Figure 5.14-1: bits 4 6 8 16 THD 9.4616e 04 5.3431e 05 3.5650e 06 4.2848e 11 1 (c)f0 = 100 bits 4 6 8 16 THD 9.1993e 04 5.5965e 05 3.0308e 06 4.5383e 11 141 1 1.5 1 0.5 --> x(n) 0.5 0 0 -0.5 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(d) As the number of bits are increased, THD is reducedconsiderably. 5.15 (a)Refer to fig 5.15-1 (b) Refer to fig 5.15-2 f=0.25 f=0.2 f=0.1 -14 f=0.5 x 10 0 50 100 0 50 100 Figure 5.15-1: The response of the system to xi(n)can be seen from fig 5.15-3 5.16 (a) 8 0 50 100 -1 -0.5 0 0.5 1 0 50 100 -1 -0.5 0 0.5 1 -1 -0.5 0 0.5 1 -4 -2 0 2 4 H(w)= h(n)e n=-8 -jwn -1 8 1 1

)-n -jwn e )n -jwn e ( +( = 3 3 = = n=-8 n=0 1 2ejw 11 3ejw + 1 11 3ejw 4 53cosw 142 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

magnitude 1 0.8 0.6 0.4 0.2 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 freq(Hz) 0 phase -0.5 -1 0 0.05 0.1 0.15 0.2 0.25 freq(Hz) 0.3 0.35 0.4 0.45 0.5 f=0.25 Figure 5.15-2: f=0.2 0.1 0.15 0.1 0.05 0.05 0 0 -0.05 -0.05 -0.1 -0.1 0 50 100 0 50 100 f=0.1 -15 f=0.5

x 10 0.3 2 0.2 0 0.1 -2 0 -4 -0.1 -0.2 -6 0 50 100 0 50 100 Figure 5.15-3: 143 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

4 |H(w)| = 53cosw H(w) =0 (b)(1) 3pn x(n)= cos 8 3pn 3pn X(w)= p d(w )+d(w +) ,-p = w = p 88 Y(w)= H(w)X(w) 4p 3pn 3pn = d(w )+d(w +) 3 88 53cos 8 3pn Hence, the output is simply y(n)= Acos 8 3p 3 where A = H(w)|= H( w= 8 8) (2)

x(n)= ..., -1,1,-1,1,-1,1,-1,1,-1,... . = cospn, -8 <n< 8 4 41 H(w)|w= == = 53cosp 82 1 y(n)= cospn 2 p Y(w)= [d(w p)+d(w + p)] 2 5.17 (a) y(n)= x(n)2cosw0x(n 1)+x(n 2) h(n)= d(n)2cosw0d(n 1)+d(n 2) (b) -jw -j2w H(w) =12cosw0e + e -jw0jw0jw) = (1ee -jw)(1eew + w0 w w0 = -4e -jwsin sin 22 = -2e -jw(cosw -

cosw0) |H(w)| =2|cosw cosw0| .|H(w)| = 0at w = w0 Refer to fig 5.17-1. (c) j2w when w0 = p/2,H(w) =1e j2/3 j/3 at w = p/3,H(p/3) =1e=1ep y(n)= |H(p/3)|3cos( n +30o 60o) 3 p =3cos( n 30o) 3 144 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

w0 = pi/3 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 0 0.5 1 1.5 2 2.5 3 --> |H(w)| --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.17-1: 5.18 (a) y(n)= x(n)x(n 4) -j4w H(w) =1e -j2w =2eej/2sin2w (c) The filter blocks the frequencyat w = Refer to fig 5.18-1. (b) p p p x(n) = cos 2 4n + cos n, H( 2 n)= 0 p p p y(n) = 2cos 4n, H( 4)= 2, 6 H( 4)= 0

2. 5.19 1 y(n)= [x(n)x(n 2)] 2 1 -j2w) H(w)= (1e 2 -jw = eej/2sinw 145 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 2 1.5 1 0.5 0 2 0 0.5 1 1.5 2 2.5 3 3.5 --> w --> theta(w) 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.18-1: p x(n) = 5+3sin( n +60o)+4sin(pn+45o) 2 p H(0) =0,H( )=1,H(p)=0 2 p y(n) =3sin( n +60o) 2 5.20 (a) y(n) = x(2n). This is a linear, time-varying system 8 Y(w) = . y(n)e -jwn n=-8 8

= . x(2n)e -jwn n=-8 w = X( 2) p = 1, |w| = p 2 = 0, 2 = |w| = p (b) y(n) = x 2(n). This is a non-linear, time-invariant system 146 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 Y(w)= X(w)* X(w) 2p Refer to fig 5.20-1. (c) Y(w) w 1/4 0 p/2 -p/2 Figure 5.20-1: y(n) =(cospn)x(n). This is a time-varying system 1 Y(w)= [pd(w p)+pd(w + p)]* X(w) 2p 1 = [X(w p)+X(w + p)] 2 3p = 0, |w| = 4 1 3p = 2, 4 = |w| = p 5.21 h(n)= . ( 1 4)n cos p 4 nu(n) .

(a) H(z) = 11 4cos 4z-1 12(1)cos z-1 +(1)2z-2 444 v 2 -1 1z 8 = v 2 z-1 + 1 z-2 14 16 (b) Yes. Refer to fig 5.21-1 v 2 (c) Poles atz = 1 v j 4e4 , zeros at z = 8 . 2 -jw 1e H(w)= v 8 . Refer to fig 5.21-2. 2 e-jw+ 1 e-j2w 1-

4 16 (d) 1 x(n) =( )n u(n) 4 1 X(z)= 11 z-1 4 147 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) z -1 + + + z -1 z -1 y(n) -2 /8 -1/16 2 /4 x(n) z -1 + + + z -1 z -1 y(n) -2 /8 -1/16 2 /4 Figure 5.21-1: Y(z) = X(z)H(z) v = 1 2 11 4z-1 + 1 2(12 8 z-1) 1v 2 4 z-1 + 1 16z-2 v v + 1+ 2v 2 2 8 z-1 12 4 z-1 + 1 16z-2 y(n) = 1 2( 1 4)n . 1+cos p 4

n + (1+ v 2)sin p 4 n . u(n) 5.22 y(n)= x(n)x(n 10) (a) -j10w H(w) =1e

-j5wj =2ee2 sin5w |H(w)| =2|sin5w|, p (w)= 5w, for sin5w> 0 2 p = 5w + p, for sin5w< 0 2 Refer to fig 5.22-1. (b) 148 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> theta(w) --> |H(w)| --> |H(w)| 1.4 1.3 1.2 1.1 1 0.9 0.8 0 0.511.52 2.533.5 --> w Figure 5.21-2: 2 1.5 1 0.5 0 --> w 2 0 0.5 1 1.5 2 2.5 3 3.5 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.22-1: 149 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim

itris G. Manolakis. ISBN 0-13-187374-1.

pp |H( )| =2, 6 H( )=0 10 10 v p p pp |H( )| =3, ( H( )= 33 36 v p pp (1) Hence, y(n) =2cos n +3 3sin( n ) 10 3152p H(0) =0,H( )=0 5 (2) Hence, y(n) =0 5.23

(a) . 1 jwndw h(n)= X(w)e 2p 3 = e 8 8 1 jwndw-jwndw e 2p -3 13pp 8 = sin n sin n pn 88 2 pp = sin

ncos n pn 84 (b) Let 2sin n 8 h1(n)= np Then, 2, |w|= 8 H1(w)=

0, 8 < |w| <p and p h(n)= h1(n)cos n 4 5.24 11 y(n)= y(n 1)+x(n)+ x(n 1) 22 Y(z)= 1 z -1Y(z)+X(z)+ 1 z -1X(z)

22 Y(z) H(z)= X(z) 1 -1 1+ z 2 = 11z-1 2 (a) 2 H(z)= 1 11z-1 2 1 h(n) =2( )n u(n)d(n) 2150 8 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) 8 H(w) = . h(n)e -jwn n=0 2 = 11 2e-jw 1 1 -jw 1+ e 2 = 11e-jw 2 = jw H(z)|z=e (c) 1 -j p 1+ e 2 H( )= -j 211 e 2 1j1 2 = 1+j1

2 2 2 -1 -j2tan =1e 1 2 pp 1 -1 Hence, y(n)= cos( n + 2tan ) 24 2 5.25 Refer to fig 5.25-1. 5.26 H(z) =(1e v j 4 z -1)(1e -j 4 z

-1) -1 -2 =12z + z v -jw -2jw H(w) =12e + e v =2e -jw(cosw 2 ) 2 v y(n)= x(n)2x(n 1)+x(n 2) p for x(n)= sin u(n) 4 y(0) = x(0)=0 v v 2 y(1) = x(1)2x(0)+x(-1) = 2 v v v

2 y(2) = x(2)2x(1)+x(0)=12 +0 =0 2 v v v v 2 2 y(3) = x(3)2x(2)+x(1)= 2+ =0 22 y(4) =0 151 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

|X(w)| for (a) |X(w)| for (b) |X(w)| for (b) 0 1 2 3 4 magnitude0 2 4 6 8 magnitude 01234 01234 |X(w)| for (c) |X(w)| for (d) 0.5 1 1.5 magnitude 0 2 4 6 8 10 12 magnitude 01234 01234 Figure 5.25-1: 5.27 k 1-z -1 (a)H(z)= Refer to fig 5.27-1. -1 . 1+0.9z (b) -jw 1e H(w)= k 1+0.9e-jw 2|sinw | 2 |H(w)| = kv

1.81+1.8cosw sinw 0.9sinw -1 -1 (w)= tan+ tan 1cosw 1+0.9cosw k 1-e -j k 2 1 (c)H(p)= = =20k =1 . k = -j 1+0.9e0.1 20 (d)y(n)= -0.9y(n 1)+ 1 [x(n)x(n 1)] 20 (e) p j( ) H( ) =0.014e6 6 p y(n) =0.028cos( n +134.2o) 6 5.28 1+bz-1 (a)H(z)= b0 -1 . Refer to fig 5.28-1. 1+az -1 1-0.6z

(b) Fora =0.5,b = -0.6,H(z)= b0 -1 . Sincethepoleisinsidethe unitcircleandthe 1+0.5z 152 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Figure 5.27-1: 54 filter is causal, it is also stable. Refer to fig 5.28-2. (c) -1 1+0.5z H(z)= b0 10.5z-1 + cosw 2 20 .|H(w)| = b 54 The maximum occurs at w =0. Hence, H(w)|w=0 = b cosw 9414 20 20 9b = 1 = 1 . b0 = 3 (d) Refer to fig 5.28-3. (e) Refer to fig 5.28-4. obviously, this is a highpass filter. By selecting b = -1, the frequency respons e of the highpass filter is improved. 5.29

A 2 |H(w)| = [1+r2 2rcos(w )][1+ r2 2rcos(w + )] d 11 = [2rsin(w )(1+ r 2 2rcos(w + )) dw |H(w)| 2 A +2rsin(w + )(1+ r 2 2rcos(w ))] 153 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form I: x(n) b 0 z -1 + + y(n) z -1 b b0 -a Direct form II : + + z -1 b-a b0 y(n) x(n) Figure 5.28-1: z-plane Figure 5.28-2: 154 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

|H(w)| phase |H(w)| 1 0.8 0.6 0.4 0.2 0 0 0.5 1 1.5 2 2.5 3 3.5 w phase 0 -0.2 -0.4 -0.6 -0.8 -1 0 0.5 1 1.5 2 2.5 3 3.5 w Figure 5.28-3: z-plane -b-0.8 Figure 5.28-4: 155 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(1+r 2)(sin(w )+ sin(w + )) (1+r 2)2sinwcos Therefore, cosw wr 5.30 y(n)= H(w)= = = |H(w)| = (w)= Refer to fig 5.30-1 5.31 (a) x(n)= X(z)= Hence, H(z)= = (b) Y(z)= = y(n)= =0 =2r [sin(w )cos(w + )+ sin(w + )cos(w )] =2rsin2w =4rsinwcosw 2 1+r = cos 2r 1+r2 . -1

= cos cos 2r 11 1 x(n)+ x(n 1)+ x(n 2) 42 4 11 1 -jw -j2w + e + e 42 4 -jw ( 1+e)2 2 -jw2w e cos 2 2w cos 2 H(w)= -w ( 1 4)n u(n)+u(-n 1) 1 -1 1 11 4z-1 +

1z-1, ROC: 4 < |z| < 1 Y(z) X(z) 1z-1 1+z-1, ROC: |z| < 1 3 4z-1 (11 4z-1)(1+z-1) 3 5 11 4z-1 + 3 5 1+z-1 3 1 3 ( )n u(n)(-1)n u(-n 1) 545156 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(w)| 1 0.8 0.6 0.4 0.2 0 -4-3-2-1 0 1 2 3 4 --> w 4 --> theta(w) 2 0 -2 -4 -4 -3 -2 -1 0 --> w 1 2 3 4 Figure 5.30-1: 5.32 y(n) H(w) = = b0x(n)+b1x(n 1)+b2x(n 2) b0 + b1e -jw + b2e -j2w (a) H( 2p 3 ) H(0) For linear phase, b0 select b0 These conditions yield b0 = b1 = b2 Hence, H(w) = = = = =

= b0 + b1e -j2 3 + b2e -j4 3 b0 + b1 + b2 = 1 b2. b2 (otherwiseb1 = 0). 1 3 1 3 e -jw(1+ 2cosw) = 0 (b) H(w)= 1 3(1+ 2cosw) (w)= . -w, for 1+ 2cosw > 0 -w + p, for 1+ 2cosw < 0 157 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Refer to fig 5.32-1. 1 0.8 0 0.5 1 1.5 2 2.5 3 3.5 --> |H(w)| 0.6 0.4 0.2 0 --> w 2 1 0 0.5 1 1.5 2 2.5 3 3.5 --> theta(w) 0 -1 -2 -3 --> w Figure 5.32-1: 5.33 (a) y(n) = 1 2M + 1 M . k=-M x(n k) H(w) = = 1 2M + 1 M . k=-M e -jwk 1 2M + 1 . 1+ 2 M . k=1 coswk . (b) y(n) H(w)

= = 1 4M x(n + M)+ 1 2M M-1 . k=-M+1 x(n k)+ 1 4M x(n M) 1 2M cosMw + 1 2M . 1+ 2 M-1 . k=1 coswk . 158 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The filterin(b)providessomewhatbettersmoothingbecauseofitssharperattenuationat the high frequencies. 5.34 28 H(z) =1+z + z + ... + z 9 1z = 1z-1 -j9w 1e H(w)= 1e-jw -j9w/2 esin9w/2 = e-jw/2 sinw/2 sin9w/2 -j4w = e sinw/2 sin9w/2 |H(w)| = || sinw/2 (w)= -4w, when sin9w/2 > 0 = -4w + p, when sin9w/2 < 0 2pk H(w) =0, at w = ,k =1,2,..., 8 9

The corresponding analog frequencies are kFs ,k =1,2,3,4, or 1kHz, 2kHz, 3kHz, 4 kHz. 9 9999 5.35 Refer to fig 5.35-1. l 1/2 Figure 5.35-1: 159 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

j3/4-j3/4 (1ez-1)(1ez-1) H(z)= G (11 z-1)2 2 H(w)= H(z)|z=ejw -j3/4) (1ej3/4)(1e H(0) = G (11)2 2l2 |H(w)| =1 . G =1 1 4 v l2 =2+ 2 1 G = v =0.073 4(2+ 2) 5.36 j-jw Hz(w) =1ree =1rcos(w .)+jrsin(w .)

(a) 2 |Hz(w)| = {[1rcos(w .)]2 +[rsin(w .)]2} 1 2 = [1+r 2 2rcos(w .)] 1 20log10|Hz(w)| = 10log10[12rcos(w .)+r 2] Hence proved. (b) imag. part -1 z(w)= tan real part rsin(w .) -1 = tan 1rcos(w .) Hence proved. (c) dz(w) tz(w)= g dw 1 [1rcos(w .)]rcos(w .)rsin(w .)(rsin(w .))

= r2sin2(w-) 1+ [1rcos(w .)]2 [1-rcos(w-)]2 r2 rcos(w .) = 1+r2 2rcos(w .) Hence proved. (d) Refer to fig 5.36-1. 5.37 1 Hp(w)= , r< 1 1reje-jw 160 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10 0 -10 -20 magnitude theta=0 1 0 -1 phase theta=0 group delay theta=0 20 10 0 -10 -5 0 5-5 0 5-5 0 5 magnitude theta=1.571 10 0 -10 -20 10 0 -10 -20 magnitude theta=3.142 1 0 -1 -5 0 5-5 0 5-5 0 5

1 0

-1 phase theta=1.571 phase theta=3.142 group delay theta=1.571 20 10 0 -10 group delay theta=3.142 20 10 0 -10 -5 0 5-5 0 5-5 0 5 Figure 5.36-1: 161 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) 1 |Hp(w)| = 2 {[1rcos(w .)]2 +[rsin(w .)]2} 1 1 = |Hz(w)| 1 |Hp= 20log10() (w)|dB |Hz(w)| = -20log10|Hz(w)| = -|Hz(w)|dB Hence proved. (b) 1rcos(w .)jrsin(w .) Hp(w)= [1rcos(w .)]2 +[rsin(w .)]2 rsin(w .) -1 p(w)= -tan 1rcos(w .) = -z(w) Hence proved. (c) dp(w) tp

g(w)= dw d(-z(w)) = dw dz(w) = dw = -tg z(w) Hence proved. 5.38 j-j-jw) Hz(w) =(1ree -jw)(1re e -j(w+)) = (1re -j(w-))(1re = A(w)B(w) (a) |Hz(w)| = |A(w)b(w)| = |A(w)||B(w)| |Hz(w)|dB = 20log10|Hz(w)| = 10log10[12rcos(w .)+r 2]+10log10[12rcos(w + .)+r 2] (b) Hz(w)= 6 A(w)+6 B(w) rsin(w .) rsin(w + .) -1 -1 = tan+ tan 1rcos(w .)1rcos(w + .) 162

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) dz(w) tg z(w)= dw = tA z (w)+tB(w) g r2 rcos(w .) r2 rcos(w + .) =+ 1+r2 2rcos(w .) 1+r2 2rcos(w + .) (d) 1 Hp(w)= Hz(w) 1 Therefore, |Hp(w)| = |Hz(w)| |Hp(w)|dB = -|Hz(w)|dB on the same lines of prob4.62 (e) Refer to fig 5.38-1. 10 5 0 -5 -10 -15 -20 magnitude theta=0 -5 0 5-5 0 5-5 0 5 magnitude theta=1.571 10 5

0 -5 -10 -15 -20 1.5 1 0.5 0 -0.5 -1 -1.5 p(w)= -z(w)and tp(w)= -tz(w) g phase theta=0 g group delay theta=0 200 150 100 50 0 1.5 1 0.5 0 -0.5 -1 -1.5 phase theta=1.571 group delay theta=1.571 200 150 100 50 0 -5 0 5-5 0 5-5 0 5 Figure 5.38-1: 163

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.39 (a) 2 (1a)2 |H1(w)|= (1acosw)2 + a2sin2w (1a)2 = 1+a2 2acosw 2 14a 1a |H1(w)|= 2 . cosw1 = 2 2a (b) 2 1a (1+cosw)2 + sin2w |H2(w)|=( )2 2 (1acosw)2 + a2sin2w (1a)2 2(1+cosw) = 2 1+a2 2acosw 12a |H2(w)|= 2 . cosw2 = 2 1+a2 By comparing the results of(a) and(b), we find that cosw2 > cosw1 and, hence w2 <w1 Therefore, the second filter has a smaller 3dB bandwidth.

5.40 h(n)= cos(w0n + ) = cosw0ncossinw0nsin use the coupled-form oscillator shown in figure 5.38 and multiply the two output s by cos and sin, respectively, and add the products, i.e., yc(n)cos+ys(n)sin= cos(w0n + ) 5.41 (a) jw0 y(n)= ey(n 1)+x(n) =(cosw0 + jsinw0)[yR(n 1)+jyI(n 1)]+x(n) yR(n 1)+jyI(n 1) = yR(n 1)cosw0 yI(n 1)sinw0 + x(n) +j[yR(n 1)sinw0 + yI(n 1)cosw0] (b)Refer to fig 5.41-1. (c) Y(z)= ejw0z -1Y(z)+1 1 = 1ejw0z-1 jnw0 y(n)= eu(n) =[cosw0n + jsinw0n]u(n) Hence, yR(n)= cosw0nu(n) yI(n)= sinw0nu(n) 164 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + + x(n) yR(n) yI(n) z z -1 -1 -sin w0 cos w0 sin w0 cos w 0 + + + x(n) yR(n) yI(n) z z -1 -1 -sin w0 cos w0 sin w0 cos w 0 Figure 5.41-1: (d) n 0 1 v 2 3 4 5 v 6 7 v 8 9 yR(n) 1 3 2 1 2v 0 1 2v 3 2 -1 3 2 1 2v 0 yI(n) 0 1 2 3 2 1 3 2 1 2 0 1 2 3 2 1 5.42

jw0 (a)poles: p1,2 = re jw0 zeros: z1,2 = e (b) For w = w0,H(w0) =0For w 6w0, the poles and zeros factors in H(w) = cancel, so that H(w)=1. Refertofig5.42-1. (c) 22 2 |1ejw0e-jw||1e-jw0e-jw| |H(w)|= G2 22 |1rejw0e-jw||1re-jw0e-jw| . . 2(1cos(w w0)) 2(1rcos(w + w0)) G2 = 1+r2 2rcos(w w0) 1+r2 2rcos(w + w0) p where w0 = . Then 3 2 d|H(w)| =0 .

w = p dw 3 2 2 |H(p)|=4G2()2 =1 1+r + r2 165 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Figure 5.42-1: 1 2) G = (1+r + r 3 (d) Refer to fig 5.42-2. (e) 22 2 |1ejw0e-jw||1e-jw0e-jw| |H(w)|= G2 22 |1rejw0e-jw||1re-jw0e-jw| In the vicinity of w = w0, we have 2 2 |1ejw0e-jw| |H(w)| G2 2 |1rejw0e-jw| . 2(1cos(w w0)) . 1 G2 == 1+r2 2rcos(w w0)2 1+r2 4G2 cos(w w0)= 2r -

4G2 1+r2 4G2 -1( w1,2 = w0 cos ) 2r 4G2 1+r2 4G2 B3dB = w1 w2 = 2cos -1( 2r 4G2 ) = 2cos -1(1( r 1 v 2 )2) . = 2 2( 1r v 2 )2 v = 2 1r 166 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + + + z -1 z -1 x(n) y(n) -2r cos w0 r2 -2cos w0 1+r+r2 3 + + + + z -1 z -1 x(n) y(n) -2r cos w0 r2 -2cos w0 1+r+r2 3 Figure 5.42-2: 5.43 For the sampling frequency Fs = 500samples/sec., the rejected frequency should b e w1 = 60 6 2p(200 4 2p()= p. The filter should have unity gain at w2 = )= p. Hence, 100255005 6 H( p) =0 25 4 and H( p) =1 5 6 6 j -jw)

H(w)= G(1e25 e 25 e -jw)(1e -j = Ge-jw[2cosw 2cos 6p ] 254 46 H( p) =2G|[cos( p)cos( p)]| =1 5 525 1 Hence, G = 2 64 cos pcos p 255 5.44 From(5.4.22) we have, -j2w 1e H(w)= b0 (1rej(w0-w))(1re-j(w0-w)) 2 2 |1e-j2w| |H(w0)|= b2 =1

0 (1r)2[(1rcos2w0)2 +(rsin2w0)2] 167 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

p(1r)2(12rcos2w0 + r2) Hence, b0 = 2|sinw0| 5.45 From a =(n +1)w0 =(n 1)w0 a+ a and cosa + cos =2cos cos , we obtain 22 cos(n +1)w0 + cos(n 1)w0 =2cosnw0cosw0 with y(n)= cosw0n, it follows that y(n +1)+ y(n 1) =2cosw0y(n)or equivalently, y(n) =2cosw0y(n 1)y(n 2) 5.46 a+ a sina + sin =2sin cos , we obtain 22 when a = nw0 and =(n 2)w0, we obtain sinnw0 + sin(n 2)w0 =2sin(n 1)w0cosw0 If y(n)= Asinw0n, then y(n) =2cosw0y(n 1)y(n 2) Initial conditions: y(-1) = -Asinw0,y(-2) = -Asin2w0 5.47 For h(n)= Acosw0nu(n) -1 1zcosw0

H(z)= A 12cosw0z-1 + z-2 Hence, y(n) =2cosw0y(n 1)y(n 2)+Ax(n)Acosw0x(n 1) For h(n)= Asinnw0u(n) z-1sinw0 H(z)= A 12cosw0z-1 + z-2 Hence, y(n) =2cosw0y(n 1)+y(n 2)+Ax(n)Asinw0x(n 1) 5.48 Refer to fig 5.48-1. y1(n)= Acosnw0u(n),y2(n)= Asinnw0u(n) 168 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + + 2r cos w0 -1 z -1 z -1 -A cos w 0 A sin w0 + + + 2r cos w0 -1 z -1 z -1 -A cos w 0 A sin w0 x(n) Figure 5.48-1: 5.49 8 (a)Replace z by z. We need 8 zeros at the frequencies w =0, 4, 2, 3,p Hence, 4 -8 1z H(z)= 1az-8 Y(z) = X(z) Hence, y(n)= ay(n 8)+x(n)x(n 8) y(n) 1 y(n)

2 4 ,e j 2 4 2 3 j j (b) Zeros at 1,e ,-1 ,e 4 3 (c) 2|cos4w| |H(w)| = v 12acos8w + a2 asin8w . -tan-1 , cos4w = 0 1-acos8w H(w)= asin8w ptan-1 , cos4w< 0 1-acos8w

Refer to fig 5.49-2. 5.50 We use Fs/L= 1cycle/day. We also choose nulls of multiples of 1 =0.071, which results in a 14 narrow passband of k0.067. Thus, M +1 =14 or, equivalently M =13 169 1 8 1 8 4 1 8 1 8 j j j Poles at ,-1. Refer to fig 5.49-1. a ,a e ,a e ,a e (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

X X X X X X X X -1 1 Unit circle X X X X X X X X -1 1 Unit circle Figure 5.49-1: magnitude of notch filter 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 --> f magnitude of a high pass filter 0.4 0.45 0.5 0 0.05 0.1 0.15 0.2 0.25 0.3 --> f Figure 5.49-2: 170 0.35 0.4 0.45 0.5 0 2 4 6 8 10 --> |H(f)| 0 2 4 6 8 10 --> |H(f)| (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.51 (a) -jw 11e a H(w)= 1ae-jw (11cosw)2 +(1sinw)2 2 aa |H(w)|= (1acosw)2 +(asinw)2 1+ a 1 2 a 2 cosw = 1+a2 2acosw 1 = for all w a2 1 Hence, |H(w)| = a For the two-pole, two-zero system, jw0e

-jw0e-jw) (11e-jw)(11e rr H(w)= (1re-jw0e-jw)(1rejw0e-jw) -jw 1 -j2w 12cosw0e+ r2 e r = r2e-j2w 12rcosw0e-jw + 1 Hence, |H(w)| = r2 1-2 -11 -2 cosw0z + z (b)H(z)= rr2 1-2rcosw0z-1+r2z-2 Hence, we need two delays and four multiplies per output point. 5.52 (a) 60 6p w0 = .2p = 200 50 6 6 H(z) =(1-

ej50 z -1)(1e -j50 z -1)b0 6p -2) -1 = b0(12cos z + z 50 H(w) =2b0e -jw(cosw cos 6p ) 506p |H(0)| =2b0(1cos )=1 251 b0 = 6 2(1cos ) 25 (b) 6 6 (1ej25 z-1)(1e-j25 z-1) H(z)= b0 6 6 (1rej25 z-1)(1re-j25 z-1) 6 2b0(1cos 25 |H(0)| =

) =1 6 r2 12rcos + 25 6 2 12rcos + r b0 = 25 6 2(1cos ) 25 171 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.53 h(n)= {h(0),h(1),h(2),h(3)} where h(0)= -h(3),h(1)= -h(2) 3ww Hence, Hr(w) =2(h(0)sin + h(1)sin ) 22p 3pp 1 Hr( ) =2h(0)sin +2h(1)sin )= 4882 3p 9p 3p Hr( ) =2h(0)sin +2h(1)sin )=1 4 88 1 1.85h(0)+0.765h(1) = 2 -0.765h(0)+1.85h(1) =1 h(1) =0.56,h(0)=0.04 5.54 (a) 3 -1 -2) (1z-1)(1+z-1)(12cos z+ z 4 H(z)= b0

2 4 (11.6cos z-1 +0.64z-2)(11.6cos z-1 +0.64z-2) 99 3 (2je-jwsinw)(2e-jw)(cosw cos ) 4 H(w)= b0 2 4 (11.6cos e-jw +0.64e-j2w)(11.6cos e-jw +0.64e-j2w) 99 3 4|sinw||cosw cos | 4 |H(w)| = b0 2 4 |11.6cos e-jw +0.64e-j2w||11.6cos e-jw +0.64e-j2w| 99 5p |H(

)| =1 . b0 =0.089 12(b)H(z)as given above. (c)Refertofig5.54-1. Thefilterdesignedisnotagood approximationof thedesiredrespo nse. 5.55 dX(w) Y(w)= e -jwX(w)+ dw (a) For x(n)= d(n),X(w)=1. dX(w) -jw Hence, = 0, and Y(w)= e dw . 1 jwndw h(n)= Y(w)e 2p . 1 jw(n-1)dw = e 2p 1 jw(n-1)| = e

2pj(n 1) sinp(n 1) = p(n 1) (b)y(n)= x(n 1)jnx(n). the system is unstable and time-variant. 172 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(f)| 1.5 1 0.5 0 --> f 4 2

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> phase 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f Figure 5.54-1: 5.56 8 H(w) = . h(n)e -jwn n=-8 = = 1, 0, 8 |w| = wc wc < |w|p G(w) = . g(n)e -jwn n=-8 8 = . h( n 2)e -jwn n=-8 8 = . h(m)e -j2wm m=-8

= H(2w) Hence, G(w)= . 1, 0, |w| = wc 2 and |w| = p wc 2 wc 2 < |w| < p wc 2 173 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.57 y(n)= x(n)x(n)* h(n)=[d(n)h(n)]* x(n)The overall system function is 1 H(z)and the frequency response is 1 H(w). Refer to fig 5.57-1. H(w) 1-H(w) 1 0 w c w 1 w c wp0 1-H(w) H(w) 1 1 w w p w 0 0wc c Figure 5.57-1: 5.58 (a) Since X(w) and Y(w) are periodic, it is observed that Y(w)= X(w p). Therefore, jn y(n)= ex(n)=(-1)nx(n) (b)x(n)=(-1)ny(n). 5.59 y(n)=0.9y(n 1)+0.1x(n) (a) 0.1

H(z)= 10.9z-1 p 0.1 Hbp(w)= H(w )= -j(w- ) 2 210.9e 0.1 = 1j0.9e-jw 174 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b)h(n)=0.1(0.9ej2 )nu(n) (c)Sincetheimpulseresponseis complex,arealinputsignalproducesa complex-valuedout put signal. For the output to be real, the bandpass filter should have a complex con jugate pole. 5.60 (a) Let g(n) = nh(n) Then, G(w) = j dH(w) dw 8 D = . n=-8 |g(n)|2 = 1 2p . - |G(w)|2dw = 1 2p . G(w)G * (w)dw = 1 2p . - . j dH(w) dw (-j) dH(w) dw *. dw But dH(w) dw = dH(w) dw + j|H(w)| d(w) dw . ej(w) Therefore, D = 1 2p . - (dH(w) dw 2 + |H(w)|2 d(w)

dw 2. dw (b) D consists of two terms, both of which are positive. For |H(w)|60, D is mini mized by = selecting (w)=0,inwhich casethesecondtermbecomes zero. 5.61 y(n)= ay(n 1)+bx(n),0 <a< 1 b H(z)= 1az-1 (a) b H(w)= 1ae-jw |b| |H(0)| = =1 1a b = (1a) (b) 2 b2 1 |H(w)|== 1+a2 2acosw 2 . 2b2 = 1+a 2 2acosw 1 cosw = 1+a 2 2(1a)2. 2a 175

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

= 1(4a 1a 2) 2a 2 4a 1a -1( w3 = cos ) 2a (c) (a 1)2 w3 = cos -1(1) 2a (a 1)2 Let f(a) =12a a2 1 Then f ' (a)= Therefore f(a)is maximum at a w3 increases as a . 0. (d) b w3 The 3-dB bandwidth increases as 5.62 y(n) H(w) |H(w)|

(w) Refer to fig 5.62-1. 5.63 (a) Y(z) H(z) 2a2 2 1a = > 0 2a2 = 1 and decreases monotonically as a . 0. Consequently, = (1a) 2 4a 1a -1( = cos ) 2a a . 0. = x(n)+ax(n M),a > 0 1+ae-jwM = = 1+2acoswM + a2 -asinwM -1

= tan 1+acoswM = 1 X(z)+z -1X(z) 2 Y(z) = X(z) 1 -1) = (1+z 2 z +1 = 2z 176 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

M=10, alpha = 0.1 --> phase --> |H(f)| 2 1.5 1 0.5 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f 1 0.5 0 -0.5 -1

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f Figure 5.62-1: Zero at z = -1 and a pole at z =0. The system is stable. (b) 1 Y(z)= -X(z)+z -1X(z). 2 Y(z) H(z)= X(z) 1 -1) = (-1+z 2z 1 = 2z Zero at z =1 and a pole at z =0. The system is stable. (c)

1 -1)3 Y(z)= (1+z 81(1+z)3 = 8 z3 Three zeros at z = -1 and three poles at z =0. The system is stable. 5.64 Y(z)= X(z)+bz-2X(z)+z -4X(z) 177 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Y(z) H(z)= X(z) -4 = 1+bz-2 + z j2w -j4w For b =1,H(w) =1+e+ e -jw = (1+2cosw)e |H(w)| = |1+2cosw| -w, 1+2cosw = 0 H(w)= pw, 1+2cosw < 0 Refer to fig 5.64-1. 0 0.5 1 1.5 2 2.5 3 --> |H(w)| 0 0.5 1 1.5 2 2.5 3 3.5 --> w 2 1 --> phase 0 0.5 1 1.5 2 2.5 3 3.5 0 -1 -2 -3 --> w Figure 5.64-1: -jw -j2w b = -1,H(w) =1e + e

-jw = (2cosw 1)e |H(w)| = |2cosw 1| -w, -1+2cosw = 0 6 H(w)= p w, -1+2cosw < 0 Refer to fig 5.64-2. 178 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.5 1 1.5 2 2.5 3 --> |H(w)| 0 0.5 1 1.5 2 2.5 3 --> |H(w)| 1 0 -1 -2 0 0.5 1 1.5 2 2.5 3 3.5 --> w Figure 5.64-2: 0 0.5 1 1.5 2 2.5 3 3.5 --> w 3 2 --> phase 5.65 y(n)= x(n)0.95x(n 6) (a) Y(z) H(z) z 6 = = = = X(z)(10.95z -6) (10.95z -6) z6 0.95 z6 0.95 z =

(0.95) 1 6 e j2k/6,k = 0,1,..., 5 6th order pole at z =0. Refer to fig 5.65-1. (b)Refer to fig 5.65-2. 6 z (c)=. r =(0.95) Hin(z)z6-0.95 1 6 . Refer to fig 5.65-3. (d)Refer to fig 5.65-4. 179 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

X r r=(0.95)1/6 X r r=(0.95)1/6 Figure 5.65-1: 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 0.5 1 1.5 2 --> |H(f)| --> f Figure 5.65-2: 180 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |H(f)| r r=(0.95)1/6X X XX X X Figure 5.65-3: 20 15 10 5 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f Figure 5.65-4: 181 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.66 (a) z-1 H(z) = 1z-1 z-2 z-1 = v v (11+ 2 5 z-1)(11- 2 5 z-1) 1v 1v = 5 v + 5 v 11+ 5 2 z-1 v 11- 5 2 z-1 1+ 5 If |z| > 1. 2 is ROC, then v v . h(n) = 1 v 5( 1+ 2 5 )n 1 v 5( 12 5 )n u(n) v v 51 5+ 1 If ROC is 2

< |z| < 2 , then v v h(n) = 1 v 5( 12 v 5 )n u(n)1 v 5( 1+ 2 5 )n u(-n 1) 51 If |z| < 1. 2 is ROC, then v v . h(n) = 1 v 5( 1+ 2 5 )n + 1 v 5( 12 5 )n u(-n 1) From H(z), the difference equation is y(n) = y(n 1)+y(n 2)+x(n 1) (b) 1 H(z)=

1e-4az-4 The difference equation is -4a y(n)= ey(n 1)+x(n) 1 H(z)=

(1e-az-1)(1ej2 e-az-1)(1+e-az-1)(1+je-az-1) 1111 4444 = +++ 1e-az-1 1je-az-1 1+e-az-1 1+je-az-1 If ROC is |z| > 1, then 1 -an h(n)= [1+(j)n +(-1)n +(-j)n]eu(n) 4 If ROC is |z| < 1, then 1 -an h(n)= [1+(j)n +(-1)n +(-j)n]eu(-n 1) 4 5.67 -4 Y(z) =1z -1 +3z -2 z -3 +6z -2)

= (1+z -1 +2z -2)(12z -1 +3z 182 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

X(z) = 1+z -1 + 2z -2 Therefore, H(z) = Y(z) X(z) = 12z -1 + 3z -2 . . h(n) = 1,-2,3 . 5.68 y(n) x(n) H(z) X(z) Y(z) Rxx(z) Hence, rxx(n) Rhh(z) Hence, rhh(n) Rxy(z) Hence, rxy(n) 1 = y(n 1)+x(n) 2 1 =( )n u(n) 4 Y(z) = X(z) 1 = 11z-1 2

1 = 11 z-1 4 1 = (11z-1)(11z-1) 42 -1) = X(z)X(z 1 = (11 z-1)(11 z) 44 -1 -4z = (11 z-1)(14z-1) 4 16 1 16 1 = 15 11 z-1 15 14z-1 4 16 1 16 = ( )n u(n)+ (4)n u(-n 1)

15415 -1) = H(z)H(z 1 = (11 z-1)(11 z) 22 -1 -2z = (11z-1)(12z-1) 2 41 41 = 311 z-1 312z-1 2 41 4 = ( )n u(n)+ (2)n u(-n 1) 323 -1) = X(z)Y(z 1 = (11 z-1)(11 z)(11 z) 442 16 1 16 1 128 1

= + + 17 12z-1 15 14z-1 105 11z-1 4 16 16 128 = (2)n u(-n 1)(4)n u(-n 1)+ ( 1715105 183 1 )n u(n) 4 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1) Ryy(z)= Y(z)Y(z 1 = (11 z-1)(11 z-1)(11 z)(11 z) 4242 64 1 128 1 64 1 128 1 = + + 21 12z-1 105 14z-1 21 11z-1 105 11z-1 24 64 128 64 1 128 1 Hence, ryy(n)= (2)n u(-n 1)(4)n u(-n 1)+ ( )n u(n)( )n u(n) 211052121054 5.69

(a) h(n)=10 ,9,-7,-8,0,5,3 . The roots(zeros) are 0.8084 j0.3370,-0.3750 j0.6074,-1.0,-0.7667 All the roots of H(z)are inside the unit circle. Hence, the system is minimum ph ase. -1 3z-2 4z-5 -6 (b)h(n)= {5,4,-3,-4,0,2,1} H(z)=5+4z-3 +2z+ zThe roots(zeros) are 0.7753 j0.2963,-0.4219 j0.5503,-0.7534 j0.1900 All the roots of H(z)are inside the unit circle. Hence, the system is minimum ph ase. 5.70 The impulse response satisfies the difference equation N akh(n k)= d(n),a0 =1 k=0 N n =0,. akh(-k)= a0h(0)=1 k=0 1 a0 = h(0) n =1,. a0h(1)+a0h(0) =0 -a0h(1) -h(1) a1 == h(0) h2(0) . . .

n = N,. a0h(N)+a1h(N 1)+... + aNh(0) . yields aN It is apparent that the coefficients {an} canbe determinedif weknowtheorderN andthevalues h(0),h(1),...,h(N). If we do not know the filter order N, we cannot determine th e {an}. 5.71 h(n)= b0d(n)+b1d(n D)+b2d(n 2D)(a)If the input to the system is x(n), the output is y(n)= b0x(n)+b1x(nD)+b2x(n2D). Hence, the output consists of x(n), which is the input signal, and the delayed signals x(n D)and x(n 2D). The latter may be thought of as echoes of x(n). (b) -jwD -j2wD H(w)= b0 + b1e + b2e 184 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

= b0 + b1coswD + b2cos2wD j(b1sinwD + b2sin2wD) 222 |H(w)| = b0 + b1 + b2 +2b1(b0 + b2)coswD +2b0b2cos2wD -1 b1sinwD + b2sin2wD (w)= -tan b0 + b1coswD + b2cos2wD (c) If|b0 + b2| << |b1|, then the dominant term is b1e-jwD and 222 |H(w)| = b0 + b1 + b2 +2b1(b0 + b2)coswD and |H(w)| has maxima and minima at w = k Dp,k =0,1,2,... (d) The phase (w)is approximately linear with a slope of -D. Refer to fig 5.71-1. --> |H(f)| --> phase 1 0.9 0.8 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f 4 2

1.2 1.1 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> f Figure 5.71-1:

5.72 (a) H(z) Hence, h(0) = = H(z)= B(z) A(z) = 1+bz1 1+az1 = 8 . n=0 h(n)z -n 1+ (ba)z -1 + (a 2 ab)z -2 + (a 2ba 3)z -3 + (a 4 a 3b)z -5 + . . . 1, 185 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

h(1) = ba, h(2) = a 2 ab, 3 h(3) = a 2ba, h(4) = a 4 a 3b (b) y(n)+ay(n 1) = x(n)+bx(n 1) For x(n)= d(n), h(n)+ah(n 1) = d(n)+bd(n 1) Multiply both sides by h(n)and sum. Then rhh(0)+arhh(1) = h(0)+bh(1) rhh(1)+arhh(0) = bh(0) rhh(2)+arhh(1) =0 rhh(3)+arhh(2) =0 By solving these equations recursively, we obtain b2 2ab +1 rhh(0) = 1a2 (ab 1)(a b) rhh(1) = 1a2 (ab 1)(a b) rhh(2) = -a 1a2 (ab 1)(a b) 2

rhh(3) = a 1a2 5.73 x(n)is a real-valued, minimum-phase sequence. The sequence y(n)must satisfy the conditions, y(0) = x(0), |y(n)| = |x(n)|, and must be minimum phase. The solution that satisfies the condition is y(n)=(-1)nx(n). The proof that y(n)is minimum phase proceeds as fol lows: -n Y(z)= y(n)z n -n = . (-1)n x(n)z n -1)n = x(n)(-z n = X(-z) This preservestheminimum phase property sinceafactor(1 az-1). (1+az-1) 5.74 Considerthesystemwithreal and evenimpulseresponse h(n)= 1 ,1, 1. andfrequency response 44 v 1 -1(12 1 H(w) =1+ cosw. Then H(z)= z z+ z + ). The system has zeros at z = -2

3. 244 We observe that the system is stable, and its frequency response is real and eve n. However, the inverse system is unstable. Therefore, the stability of the inverse system is no t guaranteed. 186 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.75 (a) g(n)= h(n)* x(n). G(w)= H(w)X(w) f(n)= h(n)* g(-n). F(w)= H(w)G(-w) Y(w)= F(-w) Then, Y(w)= H(-w)G(w) = H(-w)H(w)X(w) = H * (w)H(w)X(w) 2 = |H(w)|X(w) 2 But Ha(w)=|H(w)|is a zero-phase system. (b) G(w)= H(w)X(w) F(w)= H(w)X(-w) Y(w)= G(w)+F(-w) = H(w)X(w)+H(-w)X(w) = X(w)(H(w)+H * (-w)) =2X(w)Re(H(w)) But Hb(w)=2Re{H(w)} is a zero-phase system. 5.76 (a)Correct. The zerosoftheresultingsystemarethe combinationofthe zerosof thetwos ystems. Hence, the resulting system is minimum phase if the inividual system are minimum phase. (b) Incorrect. For example, consider the two minimum-phase systems. 112z-1 H1(z)= 11z-1 31 -1) -2(1+

z 3 and H2(z)= 11z-1 3-176z-1 Their sum is H1(z)+H2(z)= , which is not minimum phase. 11z-1 3 5.77 (a) 5 cosw 2 4 |H(w)|= 10 2 cosw 93= H(z)H(z -1)|z=e-jw 5 1(z + z-1) -1) 42 Hence, H(z)H(z = 10 1(z + z-1) 93-1 11 z 2 =

11 z-1 3 187 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) 2 2(1a2) |H(w)|= 1+a2 2acosw 2(1a2) -1) H(z)H(z = 1+a2 a(z + z-1) 2(1+a)(1a) -1) H(z)H(z = (1az-1)(1az) p2(1a2) Hence, H(z)= 1az-1 p2(1a2) or H(z)= 1az 5.78 j/2 -j/2 j/4 -j/4 -1) H(z) =(10.8ez -1)(10.8ez -1)(11.5ez -1)(11.5ez 3 -2) = (1+0.64z -2)(1-v z -1 +2.25z 2

(a)TherearefourdifferentFIR systemswith real coefficients: H1(z) = (1+0.64z -2)(1-v 3 -2) z -1 +2.25z 2 H2(z) = (1+0.64z -2)(1-v 3 -2) z -1 +2.25z 2 3 -2) H3(z) = (1+0.64z -2)(1-v z -1 +2.25z 2 3 -2) H4(z) = (1+0.64z -2)(1-v z -1 +2.25z 2 H(z)is the minimum-phase system. (b) H1(z) = 13 v 2 z -1 + 2.89z -2 1.92 v 2 z -3 + 1.44z -4 . 3 -1.92 . h1(n) = 1 . ,v 2 ,2.89, v 2 ,1.44 H2(z) = 0.64z 2 1.92 v 2 z + 2.443

v 2 z -1 + 2.25z -2 . -1.92 3 . h2(n) = 0.64, v 2 ,2.44 . ,v 2 ,2.25 H3(z) = 2.25z 2 3 v 2 z + 2.441.92 v 2 z -1 + 0.64z -2 . -3 1.92 . h3(n) = 2.25, v 2 ,2.44 . ,v 2 ,0.64 H4(z) = 1.44z 4 1.92 v 2 z 3 + 2.89z 2 3 v 2 z + 1 . -1.92 3 . h4(n) = 1.44, v 2 ,2.89,v 2 ,1 . , 188 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim

itris G. Manolakis. ISBN 0-13-187374-1.

(c) E1(n)= E2(n)= E3(n)= E4(n)=

{1,5.5,13.85,15.70,17.77} {0.64,2.48,8.44,12.94,18.0} {2.25,6.75,12.70,14.55,14.96} {1.44,3.28,11.64,16.14,17.14}

Clearly, h3(n)is minimum phase and h2(n)is maximum phase. 5.79 1 H(z)= 1+PN k=1 akz-k (a)The new system function is H ' (z)= H(.-1z) H ' (z)= 1 1+PN k=1 ak.kz-k If pk is apole of H(z), then.pk is apole of H ' (z). 1 Hence, .< is selected then |pk.| < 1 for all k and, hence the system is stable. |pmax| PN (b)y(n)= y(n k)= x(n) k=1 ak.k 5.80 (a)The impulse response is given in pr10fig 5.80-1. (b) Reverberator 1: refer to fig 5.80-2. 0 0.2 0.4 0.6 0.8 1 1.2 --> magnitude 0 500 1000 1500 2000 2500 --> n Figure 5.80-1:

189 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

impulse response for unit1 0 0.5 1 --> magnitude 0 100 200 300 400 500 600 --> n impulse response for unit2 0 0.5 1 1.5 --> magnitude 0 100 200 300 400 500 600 --> n Figure 5.80-2: Reverberator 2: refer to fig 5.80-2. (c) Unit 2 is a better reverberator. (d) For prime number ofD1,D2,D3, the reverberations of the signal in the differe nt sections do not overlap which results in the impulse response of the unit being more dense. (e) Refer to fig 5.80-3. (f)Refer to fig 5.80-4 for the delays being prime numbers. 5.81 (a)Refer to fig 5.81-1. (b) Refer to fig 5.81-2. 190 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> magnitude --> phase --> phase phase response for unit1 3 2 1 0 -1 -2 -3 01234567 --> w phase response for unit2 3 2 1 0 -1 -2 -3 01234567 --> w Figure 5.80-3: 1.2 1 0.8 0.6 0.4 0.2 0 0 500 1000 1500 2000 2500 --> n Figure 5.80-4: 191 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-20 -10 0 10 20 30 --> magnitude--> magnitude 0 0.5 1 1.5 2 2.5 3 3.5 --> w(rad) 0 --> magnitude --> phase -0.5 -1 -1.5 -2 -2.5 0 0.5 1 100 50 0 -50 1.5 2 2.5 3 3.5 --> w(rad) Figure 5.81-1: 0 0.5 1 1.5 2 2.5 3 --> w(rad) -100 4 2 0 -2 -4 0 0.5 1 1.5 2 2.5 3 3.5 --> w(rad) Figure 5.81-2: 192 --> phase (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

5.82 (a) B = 10kHz Fs = 20kHz 10k z1 = =0.5 20k 7.778k z2 = =0.3889 20k 8.889k z3 = =0.4445 20k 6.667k z4 = =0.3334 20k H(z) =(z 0.5)(z 0.3889)(z 0.4445)(z 0.3334) (b) Refer to fig 5.82-1. (c)Itsatisfiestheobjectives butthis filteris not recommendedinapractical applica tionbecause 0 -50 -100 -150 0 0.5 1 1.5 2 2.5 3 3.5 --> w(rad) 4 2

--> magnitude

0 -2 -4 0 0.5 1 1.5 2 2.5 3 3.5 --> w(rad) Figure 5.82-1: --> phase in a speech application linear phase for the filter is desired and this filter d oes not provide linear phase for all frequencies. 5.83 Refer to fig 5.83-1. Practical: 193 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

r =0.99 wr = Bandwidth = =0.0245 6 128 5 r =0.9 wr = 6 32 Bandwidth = =0.49 r =0.6 wr =0 Bandwidth =1.1536 Theoretical: r =0.99,wr = Bandwidth =2(1r)=0.02 6 r =0.9,wr = Bandwidth =2(1r)=0.2 6 For r very close to 1, the theoretical and practical values match. --> magnitude -20 -10 0 10 20 30 40 .... r = 0.99 ---- r = 0.9 __ r = 0.6 -4-3-2-1 0 1 23 4 --> w(rad) -4 -2 0 2 4 --> phase.... r = 0.99 ---- r = 0.9

__ r = 0.6 -4 -3 -2 -1 0 1 2 3 4 --> w(rad) Figure 5.83-1: 5.84 H(z) = (10.9ej0.4 z -1)(10.9e -j0.4 z -1)(11.5ej0.6 z -1)(11.5e -j0.6 z -1) H(z) = B(z) A(z) = (z 0.9ej0.4)(z 0.9e-j0.4)(z 1.5ej0.6)(z 1.5e-j0.6) z4 Let B1(z) = (z 0.9ej0.4)(z 0.9e -j0.4) B2(z) = (z 1.5ej0.6)(z 1.5e -j0.6) A(z) = z 4 194 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

B1(z)B2(z) = Hmin(z) A(z) -j0.6) (z 0.9ej0.4)(z 0.9e-j0.4)(z-1 1.5ej0.6)(z-1 1.5e = z4 B2(z) Hap(z)= B2(z-1) -j0.6) (z 1.5ej0.6)(z 1.5e = (z-1 1.5ej0.6)(z-1 1.5e-j0.6) Hap(z)hasa flatmagnituderesponse. Togeta flatmagnituderesponseforthesystem, conn ect a system which is the inverse of Hmin(z), i.e., 1 Hc(z)= Hmin(z) 4 z = (z 0.9ej0.4)(z 0.9e-j0.4)(z-1 1.5ej0.6)(z-1 1.5e-j0.6)

(b) Refer to fig 5.84-1 and fig 5.84-2. pole-zero plots for Hc(z) 901.5 0.5 130 210 60 240 270 120 300 150 330 180 0 pole-zero plots for compensated system 901.5 0.5 130 210 60 240 270 120 300 150 330 180 0 Figure 5.84-1: 195 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

mag for Hc(z) mag of compensated system mag of compensated system -15 -10 -5 0 5 10 --> magnitude -2 -1 0 1 2 --> magnitude -4-2024 -2 0 2 --> w(rad) --> w(rad) phase for Hc(z) phase of compensated system --> phase 4 1.5 1 2 0.5 0 -0.5 0 -2 -1 -4 -1.5 -4-2024 -4-2024 --> w(rad) --> w(rad) Figure 5.84-2: 196 --> phase (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 6 6.1 (a)Fourier transform of dxa(t)/dtis X a(F)= j2pFXa(F), thenFs = 2B 2 Xa (b) Fourier transform ofx(t)is (F)= Xa(F)* Xa(F), thenFs = 4B a (c) Fourier transform ofxa(2t)is X a(F)=2Xa(F/2), then Fs = 4B 11 (d) Fourier transform ofxa(t)cos(6pBt)is X a(F)= 2Xa(F +3B)+2Xa(F 3B)resulting in FL =2B and FH =4B. Hence, Fs =2B 11 (d) Fourier transform ofxa(t)cos(7pBt)is X a(F)= Xa(F +3.5B)+ Xa(F 3.5B)resulting 22 . FH in FL =5B/2 and FH =9B/2. Hence, kmax = B . =2 and Fs =2FH/kmax =9B/2 6.2 (a) Fs =1/T = 2B . A = T,Fc = B. (b) Xa(F)=0for |F|= 3B. Fs =1/T = 6B . A = T,Fc =3B. (c) Xa(F)=0for |F|= 5B. Fs =1/T = 10B . A = T,Fc =5B. 6.3

8 |k|X1 j2kt/Tp xa(t)= e(6.1) 2 k=-8 Since filter cut-off frequency, Fc =102.5, then terms with |n|/Tp >Fc will be fi ltered resulting 10 |k| 1 j2kt/Tp ya(t)= e 2 k=-10 10 |k| 1 Ya(F)= d(F k/Tp) 2 k=-10 Sampling this signal with Fs =1/T =1/0.005 =200 =20/Tp results in aliasing 8 1 . Y(F)= Xa(F nFs) 3 n=-8 8 . 9 |k|! 1 19 1 X. =

d(F k/Tp nFs)+ d(F 10/Tp nFs) 32 2 n=-8 k=-9 197 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

6.4 (a) x(n)= xa(nT)= nTe-nT ua(nT) = nTan ua(nT) -T where a = e. n Define x1(n)= aua(n). The Fourier transform of x1(n)is 8 n -j2Fn X1(F)= ae n=0 1 = 1ae-j2F Using the differentiation in frequency domain property of the Fourier transform X1(F) X(F)= Tj dF Tae-j2F = 2 (1ae-j2F) T = e(T+j2F) e-(T+j2F) 2 + (b) The Fourier transform ofxa(t)is 1 Xa(F)= (1+j2pF)2

Fig. 6.4-1(a) shows the original signal xa(t)and its spectrum Xa(F). Sampled sig nal x(n)and its spectrum X(F) are shown for Fs = 3 Hz and Fs = 1 Hz in Fig. 6.4-1(b) and Fig . 6.4-1(c), respectively. (c) Fig. 6.4-2 illustrates the reconstructed sugnal = 3 Hz and Fs =1 Hz. 8 . sin(p(t nT)/T) x a(t)= xa(nT) p(t nT)/T n=-8 6.5 The Fourier transfrom of y(t)= . t x(t)dt is -8 X(w) Y(w)= + pX(j0)d(w) jw Then, . 1 + pd(w), 0 = n = I jw H(w)= 0, otherwise 198 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1. xa(t) and its spectrum for Fs

0.4 1 0.3 -4-2 0 2 4 F(Hz) -5 0 510 t(sec) 0.5 X (F) a x(n)=x(nT) x(n)=x(nT) x (t) a aa 0.2 0.2 0.2 0.1 0 0 0.4 1 |Xa(F)| |X(F)| -4-2 0 2 4 0 -5 0 510 0.3 0.5 0.1 0 t(sec) F(Hz) 0.4

1 0 |Xa(F)| |X(F)| 0.3 0.5 0 0.1 -5 0 5 10 -4 -2 0 2 4 t(sec) F(Hz) Figure 6.4-1: 199 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x a (t) x a (t) -5 0 510 t(sec) 0.4 1 0.3 -4-2 0 2 4 F(Hz) 0.2 0.5 0.1 0 0 0.6 1 X (F) a 0 -5 0 510 t(sec) -4 -2 0 2 4 F(Hz) |Xa(F)| |X(F)| 0.4 0.2 0.5 -0.2 0 1 0 |Xa(F)| |X(F)| 0.3 0.2 0.5

0.1 -0.1 0 -5 0 5 10 -4-202 t(sec) F(Hz) Figure 6.4-2: 200 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

6.6 (a)B = F2-F1 isthebandwidthofthesignal. Basedonarbitrarybandpositioningfor first -order sampling, 2FH Fs,min = kmax where F2 kmax = . .. B (b) 8 x a(t)= xa(nT)ga(t nT) n=-8 where sinpBt ga(t)= cos2pFct pBt and Fc =(F1 + F2)/2. 6.7 . 8 j2FtdF ga(t)= Ga(F)e -8 . FL-mB . -FL 11 j2FtdF + j2FtdF

= ee 1.m+1 1.m -(FL-B) FL-mB . -FL+mB . FL+B 11 j2FtdF + j2FtdF + ee 1.-m 1.-(m+1) FL -FL+mB = A+ B + C + D 1 j2(FL-mB)t e -j2(FL+B)t A = e j2pBt(1.m+1) jB(m+1) e j2(FL-mB)t e -j2(FL+B)t = e j2pBt(ejB(m+1) e-jB(m+1)) jBm e -j2FLt ej2(FL-mB)t B = e j2pBt(ejBm e-jBm) -jBm

e j2FLt e -j2(FL-mB)t C = e j2pBt(ejBm e-jBm) -jB(m+1) e -j2(FL-mB)t ej2(FL+B)t D = e j2pBt(ejB(m+1) e-jB(m+1)) Combining A and D, and B and C, we obtain, 1 . j[2(FL+B)t-B(m+1)] -j[2(FL+B)t-B(m+1)] A+ D = e+ e pBtsin(pB(m +1)) j[2(FL-mB)t+B(m+1)] e -j[2(FL-mB)t+B(m+1)] -e 201 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

cos[2p(FL + B)t p(m +1)B]cos[2p(mB FL)t p(m +1)B] = 2pBtsin[p(m +1)B] 1 . j[2(FL-mB)t+Bm] -j[2(FL-mB)t+Bm] B + C = e+ e pBtsin(pBm)) j[2FLt-Bm] e -j[2FLt-Bm] -e cos[2p(mB FL)t pmB]cos[2pFLt pmB] = 2pBtsin(pmB) We observe that a(t)= B + C and b(t)= A+ D. Q.E.D. 6.8 1. 1, 0 = n = I gSH(n)= 0, otherwise 2. 8 -jwn GSH(w)= gSH(n)e n=-8 I -jwn = e n=0

sin[wI/2] -jw(I-1)/2 = e sin(w/2) 3. The linear interpolator is defined as 1-|n|/I, |n|= I glin[n]= 0, otherwise Taking the Fourier transform, we obtain 2 1 sin(wI/2) Glin(w)= I sin(w/2) Fig.6.8-1showsmagnitudeand phaseresponsesoftheidealinterpolator(dashed-dottedlin e), thelinearinterpolator(dashedline), andthe sample-and-holdinterpolator(solidline) . 6.9 (a) -j2F0t xa(t)= e . 8 -j2Ftdt Xa(F)= xa(t)e 0 . 8 -j2F0t -j2tdt = ee 0 . 8 = e -j2(F+F0)tdt

0 -j2(F+F0)t e |8 = 0 -j2p(F + F0) 1 Xa(F)= j2p(F + F0) 202 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 1 2 3 4 |G| 0 1 2 3 4 |G| -4-3-2-1 0 1 2 3 4 F 3 2 1 0 angle(G) -1 -2 -3 -4-3-2-1 0 1 2 3 4 F Figure 6.8-1: (b) j2F0n Fs x(n)= e 8 X(f)= x(n)e n=-8 8 -j2fn = e j2F0n -Fs -j2fn

e n=0 8 = e n=0 F0 -j2(F+Fs )n 1 = F0 -j2(F+Fs ) 1e (c) Refer to fig 6.9-1 (d) Refer to fig 6.9-2 (e)Aliasing occurs atFs =10Hz. 6.10 2 50+10 Since Fc+B = =3isaninteger,then Fs =2B =40Hz B 20 203 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> |Xa(F)| 1200 1000 800 600 400 200 0 0 200 400 600 800 1000 1200 Figure 6.9-1: Fs= 10 Fs= 20 0 2 4 6 8 10 12 --> 0 5 10 15 --> 0 5 Fs=

|X(F)|

|X(F)| 1015 0 102030 40 Fs= 100

0 10 20 30 40 --> |X(F)| 0 20 40 60 80 100 --> |X(F)| 0 20 40 60 0 50 100150 Figure 6.9-2: 204

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

6.11 Fc = 100 B = 12 B Fc + 2 r = .. B 106 = . . 12 = .8.83. =8 Fc + B B ' 2 = r 106 = 8 53 = 4 Fs =2B ' 53 = Hz 2 6.12 (a) x(n) .

X(w) x 2(n) . X(w)* X(w) The output y1(t) is basically the square of the input signal ya X(w) w (t). For the second system, X(w) * X(w) w -3p-2p-p 0 p 2p 3p -2p -p p 2p 0 2 spectrum of sampled xa (t), 2 spectrum of x (t) a 2 (i.e.), s(n) = x (nT) a 0 2B 2B -2B -2B 0 Figure 6.12-3: 2 x(t) . X(w)* X(w), the bandwidth is basically 2B. The spectrum of the sampled signal is a 205

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

given in fig 6.12-3. (b) xa(t)= cos40pt 40pn x(n)= cos 50 4pn = cos 5 y(n)= x 2(n) 4pn 2 = cos 5 118pn = + cos 22 5 112pn = + cos 22 5 11 y1(t)= + cos20pt 2 2 2 (t)= x (t) saa = cos 240pt 11 =

+ cos80pt 22 11 80pn s(n)= + cos 22 50 118pn = + cos 22 5 112pn = + cos 22 5 11 Hence, y2(t)= + cos20pt 22 For Fs = 30, 4pn x(n)= cos 3 2pn = cos 3 y(n)= x 2(n) 2pn 2 = cos

3 114pn = + cos 22 3 112pn = + cos 22 3 11 y1(t)= + cos20pt 2 2 2 sa(t)= x (t) a = cos 240pt 11 = + cos80pt 22 11 80pn s(n)= + cos 22 30 112pn =

+ cos 2 2 3 206 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11 Hence, y2(t)= + cos20pt 22 6.13 sa(t)= xa(t)+axa(t t), |a| < 1 t sa(n)= xa(n)+axa(n ) T Sa(w) w = 1+ae-j T If t T Xa(w) is an integer, then we may select H(z) = 1 1az-2 where t T = L 6.14 8 . n=-8 x 2(n) X(w) 8 . n=-8 x 2(n) = = = = 1 2p .

- |X(w)|2dw 1 T 8 . k=-8 Xa . w 2pk T . 1 T 8 . k=-8 Xa w T . , |w| = p 1 2p . 1 T2|Xa( w T )|2dw 1 T |Xa(.)|2Td. = 2pT2 T 1 T (.)|2d. = |Xa2pT T Also, Ea = . 8 x 2(t)dt a -8 = . 8 -8 |Xa(f)|2df

. F s 2 |Xa(f)|2df = -F 2 s 8 . Ea Therefore, x 2(n)= T n=-8 6.15 (a) . 8 -j2ftdt H(F)= h(t)e -8 207 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. T . 2T . 2T t -j2ftdt -j2ftdt t -j2ftdt = e +2e e T T 0 T T |{. }|{. }. {. . A B C Substituting a = -j2pf . aT 1 e1 A(F)= (aT 1)(-1) Ta2 a2 aT aT ee1

= + a Ta2 Ta2 |{z. |{z. |{z. A1 A2 A3 2 . a2T e aT B(F)= e a a3T/2 2e = sin(pfT) pf . a2T aT 1 ee C(F)= (a2T 1)(aT 1) Ta2 a2 a2Ta2Ta2T aT aT eeeee = -++ a

a Ta2 a Ta2 |{. . |{. . |{. . |{z. |{z. C1 C2 C3 C4 C5 a3T/2 e A1(F)+C1(F)= sin(pfT) pf a3T/2 e A2(F)+C3(F) = sin(pfT) Tapf aT/2 e A3(F)+C5(F)= sin(pfT) Tapf a3T/2 e C2(F)+c4(F)= sin(pfT) pf Then, -j2fT 2 esin(pfT) H(F)= T pf

(b) 6.16 (a) d(n)= x(n)ax(n 1) E[d(n)] = E[x(n)]aE[x(n 1)] =0 E[d2(n)]= s2 = E[x(n)ax(n 1)]2. d s22s2 =+ a 2aE[x(n)x(n 1)] xx 2s2 = s2 + a 2a.x(1) xx = s2(1+a 2 2a.x(1)) x .x(1) where .x(1) = s2 x .x(1) = .x(0) 208 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-6/T -5/T -4/T -3/T -2/T -1/T 0 1/T 2/T 3/T 4/T 5/T 6/T 0 T/2 T F |H| H(F) Hideal(F) -6/T -5/T -4/T -3/T -2/T -1/T 0 1/T 2/T 3/T 4/T 5/T 6/T 0 T/2 T F |H| H(F) Hideal(F) Figure 6.15-1: (b) d s2 . x s2 <s2 xd (1+a 2 2a.x(1)) 2a 2.x(1)=0 = da x x d a = .x(1) For this value of a we have s2 s2[1+.2 x (1)2.2 (1)] x x

(1)] (1).2(2)2.4 = x x s2[1.2 is always true if |.x(1)| > 0. Note also that |.x(1)|= 1. = (c) (d) x d()()(1)(2) n= xnaxnaxn12 22E[d()] E[()(1)(2)]n= xnaxnaxn12 d s2 s2 22 (1+a1 + a2 +2a1(a2 1).x(1)2a2.x(2)) = d x d s2 .x(1)[1.x(2)] . a1 = 1-

.2 =0 da1 (1) d d s2 . a2 = =0 .x da2 x x (2).2(1) 1.2(1) x x x x x d 13.2(1).2(2)+2.2 (2)+2.4 Then, s2 [1.2 6.17 x(t)= Acos2pFt 209 x (1)+.2 (1)]2

(1).x (1).x(2) = min (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

dx(t) dt = = -A(2pF)sin2pFt -2pAFsin2pFt dx(t) dt |max Hence, . = = 2pAF = . T 2pAFT = 2pAF Fs Refer to fig 6.17-1. Figure 6.17-1: 6.18 Let Pd denotethepowerspectral density of thequantization noise. Then(a) B F s Pn = Pddf B Fs 2B = Pd Fs s2 = e s2 x SQNR = 10log10 s2

e s2 xFs = 10log10 2BPd 210 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

s2 xFs = 10log10 +10log10Fs 2BPd Thus, SQNR will increase by 3dB if Fs is doubled. (b) The most efficient wayto double the sampling frequency is to use a sigma-del ta modulator. 6.19 (a) s2 e Se(F)= Fs pF |Hn(F)| =2|sin | Fs . B s2 = |Hn(F)|2Se(F)dF n -B . B pF s2 =2 0 4s2 e = Fs 4s2 e

= Fs 2s2 e = p (b) 2pB For << 1, Fs e 4sin2( ) dF Fs Fs . B 2pF (1cos )dF 0 Fs Fs 2pB [B sin ] 2pFs 2pB 2pB [ sin ] Fs Fs 2pB 2pB 12pB sin ()3 Fs Fs 6Fs 2s2 2pB 2pB 12pB e -

Therefore, s2 = [ ()3] n pFs Fs 6Fs 12B )3 = p2s2( e 3 Fs 6.20 (a) -1 1 z {[X(z)Dq(z)] Dq(z)} = Dq(z)E(z) 1z-1 1z-1 Dq(z)= z -1X(z)+(1z -1)2E(z) -1 Therefore, Hs(z)= z -1)2 and Hn(z) =(1z (b) pF |Hn(F)| =4sin2( )

Fs 2pF = 2(1cos( )) Fs 211 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) . B s2 e s2 = |Hn(F)|2 dF n -B Fs . B pF 2 [4( 0 Fs . B 32p4s2 e = F5 s 0 12B s2 e )2]2 dF Fs F4dF )5 = p4s2(

e 5 Fs 6.21 (a) 2p x(n)= cos n N xa(t)= x(n)| t n= T 2pt = cos NT Fs = cos2p( )t N Fs Therefore, F0 = N (b)N analog sinusoids can be generated. There are N possible different starting phases. 6.22 (a) h(t)= = = = (b) H(F) X(F) Y(F)

But C(F Fc)U * (-F Fc) F-1[C(F)U(F)] . 8 j2FtdF H(F)e -8 . 8 j2FtdF [c(F Fc)+c * (-F Fc)]e -8 j2Fct -j2Fct c(t)e+ c * (t)e j2Fct] 2Re[c(t)e = C(F Fc)+C * (-F Fc) = 1[U(F Fc)+U * (-F Fc)] 2 = X(F)H(F) = 1[C(F Fc)U(F Fc)+U * (-F Fc)C * (-F Fc)] 2 + 1[C(F Fc)U * (-F Fc)+U(F -

Fc)C * (-F Fc)] 2 = U(F Fc)C * (-F Fc)=0 . 8 = c(t)u(t t)dt = v(t) -8 212 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

j2Fct -j2Fct Hence, y(t)= 11 v * (t)e v(t)e+ 22 j2Fct] = Re[v(t)e 6.23 (a)Refer to fig 6.23-1. (b) Refer to fig 6.23-2. Zero Order hold: N = 32 thd = 0.1154 First Order hold, N = 32 thd=0.1152 0.4 0.6 0.8 1 ---> x(n) 0.4 0.6 0.8 1 ---> x(n) 0 20406080 0 20406080 Zero Order hold: N = 64 thd = 0.2331 First Order hold, N = 64 thd=0.2329 1 1 ---> x(n) ---> x(n) ---> x(n) ---> x(n) 0.5 0.5 0 0 0 50 100 150 0 50 100 150 Zero Order hold: N = 128 thd = 0.4686 First Order hold, N = 128 thd=0.4683 1

1 0.5 0.5 0 0 0 100 200 300 0 100 200 300 ---> n ---> n Figure 6.23-1: (c) Refer to fig 6.23-3. The first order hold interpolator performs better than the zero order interpolatorbecausethefrequency responseofthe firstorderholdismorecloser totheid eal interpolator than that of the zero order hold case. (d) Refer to fig 6.23-4. (e)Refertofig6.23-5. Higherorderinterpolatorswithmorememory or cubicsplineinterp olators would be a better choice. 213 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

---> x(n) ---> x(n) ---> x(n) ---> x(n) ---> x(n) Zero Order hold: N = 32 thd = 0.1154 First Order hold, N = 32 thd=0.1153 1.5 1.5 ---> x(n) ---> x(n) ---> x(n) 1 1 0.5 0.5 0 0 0 20406080 0 20406080 Zero Order hold: N = 64 thd = 0.2333 First Order hold, N = 64 thd=0.2332 1.5 1.5 1 0.5 1 0.5 0 0 0 50 100 150 0 50 100 150 Zero Order hold: N = 128 thd = 0.4689 First Order hold, N = 128 thd=0.4687 1.5 1.5 1 0.5 1

0.5 0 0 0 100 200 300 0 100 200 300 ---> n ---> n Figure 6.23-2: 214 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Zero Order Hold First Order Hold 10 20 30 40 50 0 0.2 0.4 0.6 0.8 1 10 20 30 40 50 0 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Zero Order Hold, filter spectrum First Order Hold, filter spectrum -1-0.5 0 0.5 1 -1-0.5 0 0.5 1 Figure 6.23-3: Zero Order Hold, interpolated output ----> |X(f)| ----> |X(f)| 50 40 30 20 10 0 0 100 200 300 400 500 600 First Order Hold, Interpolated output 50 40 30 20 10 0 0 100 200 300 400 500 600 ----> n Figure 6.23-4:

215 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Zero Order Hold, xi(n) Zero Order Hold, y(n) Zero Order Hold, y(n) 0 0.05 0.1 0.15 0.2 ---> xi(n) 0 0.05 0.1 0.15 0.2 ---> y(n) 0 10203040 0 10203040 First Order Hold, xi(n) First Order Hold, y(n) 0 0.05 0.1 0.15 0.2 ---> xi(n) 0 0.05 0.1 0.15 0.2 ---> y(n) 0 10203040 0 10203040 Figure 6.23-5: 216 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

6.24 (a)xp(t)= P8 n=-8 xa(t nTs)is a periodic signal with period Ts.Thefourier coefficientsina fourier series representation are Ts 1 2 -j2kt ck = Ts -Ts 2 xp(t)e Ts dt = 1 Ts . Ts 2 -Ts 2 8 . n=-8 xa(t nTs)e -j2kt Ts dt = 1 Ts 8 . n=-8 . Ts 2 -Ts 2 xa(t nTs)e -j2kt Ts dt = 1 Ts 8 . n=-8 . nTs+ Ts 2 nTs-Ts 2 xa(t ' )e -j2k(t ' +nTs) Ts dt ' = 1 Ts . 8 -8 xa(t ' )e -j2kt ' Ts dt ' 1 k

= Ts Xa( Ts ) 1 = Ts Xa(kdF) (b) Let Ts = t = Ts . 1, 22 w(t)= 0, otherwise If Ts = 2t, xa(t)= xp(t)w(t) Xa(F)= Xp(F)* W(F) 8 . k sinpFTs Xa(F)= ckd(F ) * Ts Ts pFTs k=-8 8 sinp(F k )Ts = Ts ck k Ts p(F )Ts k=-8 Ts 8 k sinp(F -

)Ts 1 = Xa(kdF) Ts ,Ts = k )Ts dF p(F k=-8 Ts (c) If T< 2t, there will be aliasing in every period of xp(t). Hence, xa(t) 6= x p(t)w(t) and consequently, xa(t)cannot be recovered from xp(t). (F-kF) P8 sin (d)From(b)(F)= (kdF)F Xak=-8 Xa(F-kF)

F 217 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

218 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 7 7.1 Since x(n) is real, the real part of the DFT is even, imaginary part odd. Thus, the remaining points are {0.125+j0.0518,0,0.125+j0.3018} 7.2 (a) x 2(l)= x2(l), 0 = l = N 1 = x2(l+ N), (N 1) = l =-1 3p x 2(l)= sin( l), 0 = l = 7 8 3p = sin( (l+8)), 7 = l =-1 8 3p = sin( |l|), |l|= 7 8 3 Therefore, x1(n) 8= x2(n m) x2(n) m=0

3p 3p 3p = sin( |n|)+sin( |n 1|)+... + sin( |n 3|) 88 8 = {1.25,2.55,2.55,1.25,0.25,-1.06,-1.06,0.25} (b) 3p x 2(n)= cos( n), 0 = l = 7 8 3p = -cos( n), 7 = l =-1 8 3p = [2u(n)1]cos( n), |n|= 7 8 3 1m Therefore, x1(n) 8= x 2(n m) x2(n) 4 m=0

= {0.96,0.62,-0.55,-1.06,-0.26,-0.86,0.92,-0.15} (c) 7 . -j kn for(a) X1(k)= x1(n)e 4 n=0 219 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

= {4,1j2.4142,0,1j0.4142,0,1+j0.4142,0,1+j2.4142} similarly, X2(k)= {1.4966,2.8478,-2.4142,-0.8478,-0.6682,-0.8478, -2.4142,2.8478} 8x2(n)= X1(k)X2(k) DFT of x1(n) = {5.9864,2.8478j6.8751,0,-0.8478+j0.3512,0, -0.8478j0.3512,0,2.8478+j6.8751} For sequences of part(b) X1(k)= {1.3333,1.1612j0.2493,0.9412j0.2353,0.8310j0.1248, 0.8,0.8310+j0.1248,0.9412+j0.2353,1.1612+j0.2493} X2(k)= {1.0,1.0+j2.1796,1.0j2.6131,1.0j0.6488,1.0, N 1.0+j0.6488,1.0+j2.6131,1.0j2.1796} Consequently, DFT of x1(n) 8= x2(n) X1(k)X2(k) N = {1.3333,1.7046+j2.2815,0.3263j2.6947,0.75j0.664,0.8, 0.75+j0.664,0.3263+j2.6947,1.7046j2.2815} 7.3 x (k)may be viewed as the product of X(k)with 1, 0 = k = kc,N kc = k = N 1 F(k)= 0,kc <k <N -

kc F(k) represents an ideal lowpass filter removing frequency components from (kc + 1)2 to p. N Hence x (n)is a lowpass version of x(n). 7.4 (a) 22 1 j n -j n x1(n) + = e e 2 N X1(k) = 2 [d(k1)+d(k+ 1)] also X2(k) = N 2j[d(k1)d(k+ 1)] So X3(k) = X1(k)X2(k) = N2 4j [d(k1)d(k + 1)] and x3(n) = N 2 sin( 2p N n) (b) R xy(k) = X1(k)X * 2(k)

= N2 4j [d(k1)d(k+ 1)] . rxy(n) = N 2 sin( 2p N n) 220 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) R xx(k)= X1(k)X1* (k) N2 = [d(k1)+d(k+1)] 4 N 2p . r xx(n)= cos( n) 2 N (d) Ryy(k)= X2(k)X2* (k) N2 = [d(k1)+d(k+1)] 4 N 2p . r yy(n)= cos( n) 2 N 7.5 (a) N-1 N-1 2 2 2 . 1 X

* j n -j n N + e N 4 x1(n)x2(n) = e n=0 n=0 N-1 1 X. 4 4 j n -j = e N + e N n +2 4 n=0 1 = 2N 4 N = 2 (b) N-1 N-1 . 1 X. 2 2 . 2 2 *

j n -j n -j j n NN N n eN x1(n)x2(n) = e + e e 4j n=0 n=0 N-1 1 X. 4 4 j -j n N n e N = e 4j n=0 =0 * (c)PN-1 x1(n)x2(n)=1+1 =2 n=0 7.6 -j 4 N-1

n n n n w(n) =0.420.25 ej 2 N-1+ e -j 2 N-1+0.04 ej 4 N-1+ e N-1 "N-1 N-1 . 2 . 2 . 2 . n -j n -j -j nk nk nk 0.25 N N + N w(k) =0.42 e ej 2 N-1e e -j 2 N-1e n=0 n=0 n=0 "N-1 N-1 . 2 .

2 n -j n -j nk nk N + N +0.04 ej 4 N-1e e -j 4 N-1e n=0 n=0 221 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

=0.42Nd(k) . j2[ N N-1 +k]# -k] -j2[ 1e 1e N N-1 -0.25 + k k j2[N-1 1 -j2[N-1 1 + N ] . j2[ 1e N ]1e 2N N-1 +k]# -k] -j2[ 1e 1e 2N N-1 +0.04 + j2[N-1 2 k ] -j2[N-1 2 + k ] N N

1e 1e =0.42Nd(k) k 1cos(2N )cos(2p( 1 + ))+cos(2k ) N-1N-1 NN -0.25 k 1cos(2p( 1 + )) N-1 N k 1cos(4N )cos(2p( 2 + ))+cos(2k ) N-1N-1 NN +0.04 k 1cos(2p( 2 + )) N-1 N 7.7 N-1 . 2k0n 2k0n . 1

j-j N Xc(k)= x(n) eN + ee -2kn N 2 n=0 N-1 N-1 1 . 2(k-k0)n 1 . 2(k+k0)n -j -j N N = x(n)e + x(n)e 22 n=0 n=0 11 = X(kk0)modN + X(k 2 2+ k0)modN 11 similarly, Xs(k)= X(kk0)modN X(k + k0)modN 2j 2j

7.8 x2(n) y(n)= x1(n) 4 3 = x1(m)mod4x2(n m)mod4 m=0 = {17,19,22,19} 7.9 X1(k)= {7,-2j,1,-2+j} X2(k)= {11,2j,1,2+j} . X3(k)= X1(k)X2(k) = {17,19,22,19} 222 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

7.10 x(n) x(n)x * (n) E 7.11 (a) (b) 7.12 (a) (b) s(k) s(n) s(n) = = = = = = 2kn N 1 . j -j = e + e 2kn N 2 1 . 4kn N j -j = 2+e + e 4kn N 4 N-1

= x(n)x * (n) n=0 N-1 4kn N 1 X. j -j = 2+e + e 4kn N 4 n=0 1 = 2N 4 N = 2 x1(n)= x(n 5)mod8 -j X1(k)= X(k)e 25k 8 5k -j 4 = X(k)e

x2(n)= x(n 2)mod8 -j X2(k)= X(k)e 22k 8 k -j 2 = X(k)e Wk 2 X(k) (-1)kX(k) 5 1 . (-1)kX(k)W-kn N 6 k=0 5 1 . -k(n-3) X(k)W N 6 k=0 x(n 3)mod6 {3,4,0,0,1,2} N =6 X(k)+X*(k) y(n) = IDFT 2 223 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 = [IDFT{X(k)} +IDFT{X * (k)}] 2 1 = . x(n)+x * (-n)modN 2 . x(1)+x(5) x(2)+x(4) x(4)+x(2) x(5)+x(1) . = x(0), , ,x(3),, 22 22 11 =0, ,3,3,3, 22 (c) X(k)X*(k) v(n) = IDFT 2j By similar meansto(b) 11 v(n) =0,j,j,0,-j, j 22 7.13 (a) N-1

x(n)Wkn X1(k)= N n=0 3N-1 x(n)Wkn X3(k)= 3N n=0 N-12N-13N-1 x(n)Wkn x(n)Wkn x(n)Wkn = 3N + 3N + 3N n=0 n=Nn=2N N-1 N-1 N-1 k 3 k 3 k 3 n n n x(n)W3 kW x(n)W2k 3 = x(n)W W + + N N

N n=0 n=0 n=0 N-1 . n = x(n)1+W3 k + W2k W 3 N n=0 = (1+W3 k + W2k)X1(k) 3 (b) X1(k) =2+Wk 2 k 3 X3(k) =2+W6 k +2W2k + W3k +2W4k + W5k 66 66 k 3 k 3 k 3 )+W2k 6 (2+W )+W4k 6 = (2+W (2+W ) 2

2 2 = (1+Wk 3 + W2k 3 )X1( k 3) 7.14 (a) y(n) = = x2(n) x1(n) 5 {4,0,1,2,3} 224 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) Letx3(n)= {x0,x1,...,x4}. Then, . .. . . x0 . 04321 1 10432 21043 32104 0 0 0 .... .... .... x1 x2 x3 .... = .... .... 43210 x4 0 Solving yields sequence

x3(n)= -0.18 ,0.22,0.02,0.02,0.02 . . 7.15 . Define H1(z)= H-1(z) and corresponding time signal h1(n). The use of 64-pt DFTs of y(n) y(n) 64 h1(n)whereas x(n)requires linear convolution. However we and h1(n)yields x(n)= can simply recognize that X(z)= Y(z)H1(z) -1 = Y(z)0.5Y(z)z so x(n)= y(n)0.5y(n 1), 0 = n = 63 . with y(-1) =0 7.16 N-1 H(k)= h(n)e 2 -j kn N n=0 2 k0k 1 -j = 1+( )e 4k0

41

-j k 2 =1e 4 1 G(k)= H(k) 1 = -j k N 11 e2 4 2 1 = 1+ e -j 1

k -j k

+ + ... 2 2 e 44 4 164j 4 16+4j = ,, ,, repeat k0 times 317 517 N-1 1 2 j kn N g(n)= G(k)e N n=0 4k0-44k0-3 164j 1 2 4 kn 2 j

j kn [ 4k0 4k0 + = e e 4k0 3 17 k=0,4,... k=1,5,... 225 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

4k0-24k0-1 2 2 kn X16+4j j . 4 j . kn 4k04k0 + e+ e] 5 17 k=2,6,... k=3,7,... 4 . 4 . 2 2 1 X164j j 2n X n Xj 4k0 4k0 = + e+ e 4k0 317 5 2 16+4j j 3n X 4k0 + e 17 k0-1 . . .

j2 ni k0 where = e i=0 But =1, yielding 4 . . 1 164j 4 16+4j g(0) = ++ + 43 17 5 17 256 = 255 1 4 164j 4 16+4j g(k0)= + j j 43 17 5 17 64 = 255 4 . . 1 164j 4 16+4j g(2k0)= + -

43 17 5 17 16 = 255 4 . . 1 164j 4 16+4j g(3k0)= j + j 43 17 5 17 4 = 255 and g(n) = 0 for other n in [0,4k0). .. .. .. .. .256 1 . Therefore, g(n)* h(n)= ,0,0,..., 0,..., 0 ,..., 0 ,..., ,0 255.. . 255 .. .. . .k0 2k0 3k0 . .. 4k0

g(.)represents a close approximation to an inverse system, but not an exact one. 7.17 7 . 2 -j kn X(k)= x(n)e 8 n=0 = {6,-0.7071j1.7071,1j,0.7071+j0.2929,0,0.7071j0.2929,1+j, -0.7071+j1.7071} |X(k)| = {6,1.8478,1.4142,0.7654,0,0.7654,1.4142,1.8478} -pp X(k) =0,-1.9635, ,0.3927,0,-0.3927, ,1.9635 44 226 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

7.18 8 x(n)= d(n iN) i=-8 y(n)= h(m)x(n m) m = h(m) d(n m iN) mi = h(n iN) i Therefore, y(.)is a periodic sequence with period N. So N-1 Y(k) = . y(n)Wkn N n=0 = H(w)|w= 2 N k 2pk Y(k) = H( N ) k = 0,1,. . . ,N 1 7.19 Call the two real even sequences xe1(.)and xe2(.), and the odd onesxo1(.)and xo2 (.)(a) Let xc(n) =[xe1(n)+xo1(n)]+j[xe2(n)+xo2(n)] Then, Xc(k) = DFT{xe1(n)} +DFT{xo1(n)} + jDFT{xe2(n)} + jDFT{xo2(n)} =[Xe1(k)+Xo1(k)]+j[Xe2(k)+Xo2(k)] Re[Xc(k)]+Re[Xc(-k)] where Xe1(k)= 2

Re[Xc(k)]Re[Xc(-k)] Xo1(k)= 2 Im[Xc(k)]+Im[Xc(-k)] Xe2(k)= 2 Im[Xc(k)]Im[Xc(-k)] Xo2(k)= 2 (b) si(0) = xi(1)xi(N 1) =0 -si(N n)= -xi(N n +1)+ xi(N n 1) = xi(n +1) xi(n 1) = si(n) (c) x(n) =[x1(n)+s3(n)]+j[x2(n)+s4(n)] TheDFT ofthefoursequences canbe computed usingtheresultsofpart(a) N-1 si(n)Wkn For i =3,4,si(k)= N n=0 227 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N-1 =[xi(n +1) xi(n 1)]Wkn N n=0 W-k = Xi(k)WN k Xi(k) N 2p =2jsin( k)Xi(k) N s3(k) Therefore, X3(k)= 2jsin(2k) N s4(k) X4(k)= 2jsin(2k) N (d)X3(0)and X4(0), becausesin(2k)=0. N 7.20 N-1 x(n)Wkn X(k)= N n=0 N 2 -1

N 2 -1 N N 2 ) k(n+ x(n)Wkn N + x(n + )W = N 2 n=0 n=0 N 2 -1 X. . Wkn = x(n)x(n)Wk 2 N n=0 If k is even, Wk =1, and X(k)=0 2 (b) Ifk is odd, Wk = -1, Therefore, 2 N 2 -1 2x(n)Wkn

X(k)= N n=0 N 2 -1 . n =2 x(n)W n=0 N 2 k 2 N For k =2l+1,l =0,..., 1 2 N 2 -1 N 2 x(n)Wln n=0 N n = pt DFT of sequence 2x(n)W N 2 7.21

(a)Fs = FN =2B =6000 samples/sec (b) 1 T = Fs 1 = 6000 228 Wn N X(2l+1) =2 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 LT . L = = = = 50 1 50T 6000 50 120 samples (c)LT = 1 6000 120 = 0.02 seconds. 7.22 1 1 2 22 N-1 N 2 kn N j -j e n n , 0 = n = N, N =10 x(n) +

= e 2 N-1 N-1 N -j X(k)= x(n)e n=0 1 1 N 22 22 n=0 n=0 =5d(k1)+5d(k9), 0 = k = 9 7.23

N -j (k-1)n -j (k+1)n + = e e PN-1

2 (b) N-1 N -j kn (a)X(k) d(n)e 1, 0 = k = N 1 = = n=0 N N N 2 kn0 , 0 = k = N 1 2 2 -j kn N X(k)= d(n n0)e

n=0 2 -j = e (c) N-1 -j kn n X(k)= a e n=0 N-1 -j k)n =(ae n=0 N N 1a = 2 -j k 1ae (d) N -1 2 X(k) = e 2

N-j n=0 = 2 N -j2 N -j1e1e kn N k 2 k N 1(-1)k = 2 -j k 1e229 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(e) N-1 2 2 -j e N N 2 (k-k0)n N j e nk0-j kn X(k)= e n=0 N-1 = n=0 = Nd(kk0) (f) 1 1 2 N N

From(e) we obtain X(k)= [d(kk0)+d(kN + k0)] 2 (g) 2 2 22 N N j e nk0 + -j nk0 x(n)= e 1 1 2 2j 2j N N j nk0 -j nk0 x(n)= e

e N Hence X(k)= [d(kk0)d(kN + k0)] 2j (h) 2 N-1 X -j nk( X(k)= x(n)e N N assumeN odd) 22 n=0 N N 2 2 -j -j 2k -j + 4k -j + ... +

(n-1)k 1+ = e e e N+1 2 2k) 1(e= -j1e2 N 2-jk N N N 4 2 2k 1e = -j k 1e 1 = 1e 7.24 (a) -j

k N-1 1 N 2 -2j j nk x(n)= X(k)e N k=0 N-1 N 2 k=0 X(0)+X(1)+X(2)+X(3) =4 j nk . X(k)e = Nx(n) 3 X(0)+X(1)e j

2 + X(2)e j j + X(3)e

2 =8 j j2j3 X(0)+X(1)e+ X(2)e+ X(3)e= 12 3 9

X(0)+X(1)e j j3j + X(3)e + X(2)e 4 2 2 = . . . .. .. .. . .. . 1111 X(0) 4 X(0) 7

1 j -1 -j 1 -11 -1 X(1) X(2) 8 12 X(1) X(2) ... .. . .. . .. . = .. . .. . . .. . .. . = ... 1 1 -j -1 jX(3) 4 X(3) -2+j 230 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) 3 X(k)= x(n)e 2 -j nk 4 n=0 3 X(0) = x(n) n=0 = 7 3 X(1) = x(n)e

-j n 2 n=0 = -2j 3 X(2) = x(n)e n=0 = 1 3 -jn X(3) = x(n)e 3 -j n 2 = n=0 -2+j 7.25

(a) 8 . X(w)= x(n)e n=-8 -jwn -jw -j2w = ej2w +2ejw +3+2e + e = 3+2cos(2w)+4cos(4w) (b) 5 V(k)= v(n)e 2 -j nk 6 n=0 2 2 2 2 -j k -j -j 4k -j 5k 66 66 = 3+2e + e 2k +0+ e + e p 2p = 3+4cos( k)+2cos( k) 33

(c)V(k)= X(w)|k 2k w= = 63 This is apparent from the fact that v(n)is one period(0 = n = 7) of a periodic sequence obtained by repeating x(n). 7.26 8 Let x(n)= d(n + lN) l=-8 Hence, x(n)is periodic with period N, i.e. x(n) =1,n =0,N,2N,. . . 231 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

=0, N-1 2 2 2 otherwise N N N -j nk Then X(k)= x(n)e 1, 0 = k = N 1 = n=0 N-1 1 . j X(k)e nk and x(n) = N k=0 8 N-1 . 1 .

j nk Hence, d(n + lN)= e N l=-8 k=0 7.27 (a) M-1 y(n)Wkn M Y(k)= M n=0 M-1 x(n + lM)Wkn = M n=0 l -jwn Now X(w)= x(n)e, n 2p . Wkn so X( k)= M M n M-1 X. k(n+lM)

= x(n + lM)W M n=0 l M-1 x(n + lM)Wkn = M n=0 l = Y(k) 2 Therefore, Y(k)= X(w)| w= k (b) Y(k)= X(w)| w= N 2 2 k k Y( )= X(w)| 2 N 2 = X(k),k =2,4,...,N 2 (c) X1(k)= X(k+1) k w=

N 2 = x(n)Wn N N Let y(n)= x1(n)+x1(n + ), 0 = n = N 1 2 =0, elsewhere Then X(k+1) = X1(k) k = Y( ),k =0,2,...,N 2 2 232 -j n . x1(n)= x(n)e (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N where Y(k)is the -pt DFT of y(n) 2 7.28 (a)Refer to fig 7.28-1. (b) x(n) X(w) 1 20 0.8 15 ---> X(w) ---> x(n) 0.6 10 0.4 5 0.2 0 0 -5 0 10 20 30 01234 ---> n ---> w ck xtilde(n) 0.6 ---> xtilde(n) 1 0 10 20 30 ---> n

0.4 0.2 0.5 0 -0.2 0 0 10 20 30 Figure 7.28-1: 88 x(n)e -jwn = a |n|-jwn e n=-8 n=-8 1 L -n -jwn n -jwn a + + = ae ae 1 -L LL jwn + -jwn n

n ae a + = ae 11 L n a +2 cos(wn) = a n=1 L = x(0)+ 2 n=1 x(n)cos(wn) (c) Refer to fig 7.28-1. (d) Refer to fig 7.28-1. 233 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(e) Refer to fig 7.28-2. (f)N=15. Refer to fig 7.28-3. x(n) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 ---> x(n) 0 50 100 150 200 250 ---> n Figure 7.28-2: 234 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) X(w) 1 20 0.8 15 ---> X(w) 0.6 10 0.4 5 ---> ck ---> x(n) 0.2 0 0 -5 -20-1001020 0 1 2 3 4 ---> n ---> w ck xtilde(n) 1.5 1.4 ---> xtilde(n) 1 0.5 1.2 1 0.8 0 -0.5 0.6

0 5 10 15 -10 -5 0 5 10 ---> w ---> n Figure 7.28-3: 7.29 Refer to fig 7.29-1. The time domain aliasing is clearly evident when N=20. 7.30 Refer to fig 7.30-1. (e) xam(n)= x(n)cos(2pfcn) N-1 -j2fn Xam(w)= x(n)cos(2pfcn)e n=0 N-1 1 X. i -j2(f-fc)n -j2(f+fc)n = x(n) e + e 2 n=0 1 Xam(w)= [X(w wc)+X(w + wc)] 2 7.31 2 { 2 ,1 (a)ck = ,

,1 ...} 32 (b) Refer to fig 7.31-1. The DFT of x(n)with N =128 has a better resolution comp ared to one with N =64. 235 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> mag --> mag --> mag --> mag --> mag X(w) x(n) 10 1.5 1 5 0.5

--> x(n) --> x(n) --> x(n) 0 0 0 2 4 6 8 0 100020003000 X(w) with N=20 x(n) with N=20 1.5 1 0.5 5 0 0 02468 0 10 20 X(w) with N=100 x(n) with N=100 10 1.5 1 0.5

5 0 0 02468 0 50 100 --> w --> n Figure 7.29-1: 7.32 (a) 1 Y(j ) = P(j )* X(j ) 2p 1 T0 2 T0 -j T0sin( )e * [2pd( 0)] = 2p 2 . sin x where sincx = x T0( 0)

T0( 0) Y(j ) = T0sinc 2 e -j2 (b)w0P = 2pk for an integer k, or w0 = 2k P p (c) N-1 Y(w) = . ejw0n e -jwn n=0 236 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) xc(n) -2 0 2 x(n) -1 0 1 xc(n) 0 100 200 300 0 100 200 300 xam(n) Xam(w) with N=128 -2 0 2 xam(n) 0 20 40 mag 0 100 200 300 0 50 100 150 Xam(w) with N=100 Xam(w) with N=180 30 60 20 40 0 50 100 150 0 100 200 300 mag mag 10 20 0 0 Figure 7.30-1: = e sinw-w0 2

sinN 2 (w w0) -j N-1 2 (w-w0) Larger N . narrower main lobe of |Y(w)|. T0 in Y(j ) has the same effect. (d) Y(k)= Y(w)|2 w= k N N-1 sinp(kl) -j (k-l) N = e (k-l) sin N |sinp(k l)| |Y(k)| = (k-l) |sin| N = Nd(kl) 2 (e) The frequency samples

k fall on the zeros of Y(w). By increasing the sampling by a factor N of two, for example, we will obtain a frequency sample between the nulls. Y(w)| w= 2 k, k=0,1,...,2N-1 k= 2NN 237 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) DFT of x(n) with N=64 DFT of x(n) with N=64 0 200 400 600 800 -1.5 -1 -0.5 0 0.5 1 1.5 0 20 40 60 80 0 5 10 15 20 DFT of x(n) with N=128 0 50 100 150 0 10 20 30 40 50 60 Figure 7.31-1: 238 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 8 k is an Nth 2 of unity. Consider PN-1 n=0 N 2 knj e N N 2 ln . If k 6= l, the terms in the sum represent the N equally j j2k j k)N = e=1,e k satisfies the equation XN = 1. Hence ej Since (e root 22 spaced roots in the unit circle which clearly add to zero. However, if k = l, th e sum becomes PN-1 1= N. see fig 8.1-1 n=0 N N

j e p 12e p 12e unit circle z-plane j4 j2 Roots for N=12 Figure 8.1-1: q(l-1) N N 22 (b)Let W q Wq +d where W q N = N N 2 is the truncated value ofWq ql -j -j q(l-1) -j = e ql Wq N q (a) W W

= e e = N N ql ql N. Now W = +d)l W (Wq N +ld. N N N 239 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Generally, single precision means a 32-bit length or d =5x10-10; while 4 signifi cant digits means d =5x10-5 . Thus the error in the final results would be 105 times larger. ql (c) Since the error grows as ld, after N iterations we have an error of Nd. If W is reset to -j N after every ql = NN d. Thus, the iterations, the error at the last step of the iteration is ld = 44q error reduced by approximately a factor 4q. 8.3 N-1 x(n)Wkn X(k)= 0 = k = N 1 N n=0 N 2 -1 N-1 x(n)Wkn x(n)Wkn =+ NN n=0 n= N 2 N 2

-1 N 2 -1 . N N 2 (r+ )k x(n)Wkn N + x(r + )W = N 2 n=0 r=0 ' (k ' )0 = k ' N LetX = X(2k+1), = 1 2 N 2 -1 N )(2k ' N 2 (2k ' (n+ +1)

+1)n Then, X ' (k ' ) = x(n)W + x(n + )W 2 N N n=0 Using the fact that W2k ' N N 2 n Wk ' n WN N = 1 = , N 2 -1 N N 2 N 2 )Wk ' n X

' (k ' ) N 2 x(n)WN n Wk ' n nn NW + x(n + W = N 2 n=0 N 2 -1 . N N 2 WN n Wk ' = x(n)x(n + ) 2 n=0 8.4 Create three subsequences of 8-pts each N 3

N 3 N 3 21 22 23 y(n)Wkn y(n)Wkn y(n)Wkn Y(k)= ++ N NN n=0,3,6,... n=1,4,7,... n=2,5,... 77 7 y(3i)Wki y(3i+1)Wki y(3i+2)Wki Wk N + W2k N + = i=0 i=0 i=0 . = Y1(k)+WN k Y2(k)+WN 2kY3(k) where Y1,Y2,Y3 represent the 8-pt DFTs of the subsequences. 8.5 -1 -6 X(z) =1+z + ... + z 240 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

X(k)= X(z)|2 j z=e5 2 4 12 -j -j -j 55 5 = 1+e + e + ... + e 2 4 8 -j -j -j 55 5 = 2+2e + e + ... + e x ' (n)= {2,2,1,1,1} x ' (n)= x(n +7m),n =0,1,..., 4 m Temporal aliasingoccursin firsttwopointsof x ' (n)because X(z)is not sampled at sufficiently small spacing on the unit circle. j[2k + 88 (a)Zk =0.8e ] see fig 8.6-1 (b) circle of radius 0.8 z1 z0

z2 z3 z4 z5 z6 z7 2 p 8 p 8 z-plane Figure 8.6-1: X(k) s(n) = = = X(z)|z=zk 7 . n=0 x(n) . 0.8ej[2k 8 + 8 ]i-n x(n)0.8e -j 8 n 241 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8.7 N Let M = ,L =2. Then 2 N 2 -1 x(0,m)Wmq F(1,q)= x(1,m)Wmq n=0 which are the same as F1(k)and F2(k)in(8.1.26) N 2 N 2 F(0,q)= n=0 N 2 -1 G(0,q)= F(0,q)= F1(k) G(1,q)= Wq F(1,q)= F2(k)Wk NN X(0,q)= x(k)= G(0,q)+G(1,q)W0 2 = F1(k)+F2(k)Wk N X1,q)= x(k)= G(0,q)+G(1,q)W1 2 = F1(k)F2(k)Wk

N 8.8 1 W8 = v (1j) 2 Refer toFig.8.1.9. The firststageof butterflies produces(2,2,2,2,0,0,0,0). Thetw iddle factormultiplicationsdo notchangethissequence. The nex stageproduces(4,4,0,0,0,0 ,0, 0) which againremains unchangedby thetwiddlefactors. Thelaststageproduces(8,0,0, 0,0, 0, 0, 0). The bit reversal to permute the sequence into proper order unscrambles only zeros so the result remains(8,0,0,0,0,0,0,0). 8.9 See Fig. 8.1.13. 8.10 Using(8.1.45),(8.1.46),and(8.1.47) thefig 8.10-1is derived: 8.11 Using DIT following fig 8.1.6: 111 1. 1st stage outputs : , , ,..., 2222 1111 2nd stage outputs : 1, (1+W82),0,

(1W82),1, (1+W82),0, (1W82) 2222 242 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(0) x(0) x(4) x(1) x(8) x(2) x(12) x(3) x(1) x(4) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 0 2 4 6 0 3 6 9 x(5) x(9) x(13) x(2) x(6) x(10)

x(14) x(3) x(7) x(5) x(6) x(7) x(8) x(9) x(10) x(11) x(12) x(13) x(11) x(14) x(15) x(15) Figure 8.10-1: . 2, 1 1 3rd (1+W 18 28 38 28 38 + W + W ),0,

(1W + W W ),0, stage outputs : 2 2 1 1 18 + W 28 38 28 38 58 (1W W ),0, (1W W + W 2 2 Using DIF following fig 8.1.11: 1111111 1 2 W

1 2 1st 18 , 28 38 W W stage outputs : , , , , , , , 2 2 2 2 2 2 2 1 1 ),0, (W 2 1

1 2nd stage outputs : 1,1,0,0, 28 18 38 28 (1+W + W ), (1W ), (W W ) 2 2 2 1 1 3rd stage outputs : 2,0,0,0, 18 28 38 18 28 (1+W + W + W ),

(1W + W W ), 2 2 1 2 1 (1W 28 38 W 58 28 38 + W 58 + W ), (1W W 2 8.12 Let . . 1111

1 -j -1 j 1 -11 -1 1 j -1 -j . A = .. . .. . . x(0) x(4) x(8) x(12) T x1 = 243 58 58 38 38 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. x(1) x(5) x(9) x(13) T x2 = . x3 = x(2) x(6) x(10) x(14) T . x4 = x(3) x(7) x(11) x(15) T . . . . F(0) 4 F(4) F(8) 0 0 .. .. = Ax1 = .. .. F(12) 0 . . .

. F(1) 0 F(5) F(9) 0 0 .. . .. . = Ax2 = .. . .. . F(13) 0 . . . . F(2) -4 F(6) F(10) 0 0 .. . ..

. = Ax3 = .. . .. . F(14) 0 . . . . F(3) 0 F(7) F(11) 0 0 .. . .. . = Ax4 = .. . .. . F(15) 0 As every F(i)=0 except F(0)= -F(2)=4, . .

. .. . x(0) F(0) 0 x(7) x(8) F(1) F(2) 8 0 .. . .. . = Ax4 .. . ... = .. . .. . x(12) F(3) 8 which means that X(4)= X(12)=8.X(k)=0forother K. 8.13 (a) gain = W80W80(-1)W82 = -W82 = j

(b) Given a certain output sample, there is one path from everyinput leading to it. This is true for every output. (c) X(3) = x(0)+ W83x(1)W82x(2)+ W82W83x(3)W80x(4)W80W83x(5)+ W80W82x(6)+ W80W82W83x(7) 8.14 Flowgraph for DIF SRFFT algorithm for N=16 is given in fig 8.14-1. There are 20 real, non trivial multiplications. 244 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(0) X(0) x(1) X(8) x(2) X(4) x(3) X(12) x(4) X(2) x(5) x(6) x(7) x(8) x(9) x(10) x(11) x(12) x(13) x(14) x(15) -1 -1 -1 -1 -1 -1 -1 -1 +j +j +j +j w 0 w 1 w 2 w3 w

w w w 0 3 6 9 -1 -1 -1 +j -1 -1 -1 +j -1 +j -1 w 0 w 2 w0 w 6 +j +j -1 -1 -1 -1 -1 +j -1 -1 -j -j -j -j -j -j -j -j -j -j Figure 8.14-1: 8.15 For the DIT FFT, we have X(10) X(6) X(14) X(1) X(9) X(5) X(13)

X(3) X(11) X(7) X(15) N 2 -1 N 2 -1 N 2 X. (2n+1)k x(2n)Wnk X(k) = + x(2n +1)W N n=0 n=0 The first term can be obtained from an N -point DFT without any additional multi plications. 2 Hence, we use a radix-2 FFT. For the second term, we use a radix-4 FFT. Thus, fo r N=8, the DFT is decomposed into a 4-point, radix-2 DFT and a 4-point radix-4 DFT. The lat ter is N 2 -1 N 4 -1 N 4 -1

N 4 . (2n+1)k X. NWk x(4n +1)Wk 3kk(4+3)WWxnNN 4 x(2n +1)W + = N n=0 n=0 n=0 NN 3NN The computation of X(k),X(k+ ),X(k+ ),X(k+ )fork =0,1,..., 1 areperformed 424 4 from the following: N 2 -1 N 4 -1 N 4 -1 N 2 x(2n)Wnk

N 4 NWnk x(4n +1)Wk N 4 N Wnk x(4n +3)W3k X(k) + + = n=0 n=0 n=0 N 2 -1 N 4 -1 N 4 -1 N N 4 N 2 x(2n)Wnk x(4n +1)(-j)Wnk N 4 N (j)Wnk x(4n +3)W3k (-1)n X(k+

) + + = 4 n=0 n=0 n=0 N 2 -1 N 4 -1 N 4 -1 N N 4 N 2 x(2n)Wnk NWnk x(4n +1)(-1)Wk N 4 N Wnk x(4n +3)(-1)W3k X(k+ ) + + = 2

n=0 n=0 n=0 245 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N 2 -1 N 4 -1 N 4 -1 3N x(2n)Wnk NWnk N 4 N 2 N 4 N Wnk x(4n +3)(-j)W3k (-1)n n X(k+ ) x(4n +1)(j)W + + = 4 n=0 n=0 n=0 The basic butterfly is given in fig 8.15-1 W3k WN n from the use of x(2n)

-1 -1j -j X X X(k) X(k+N/4) Note that this is a mirror image of DIF-SRFFT butterfly. from x(2n+1) from X(k+N/2) X(k+3N/4) x(4n+3) N x(0) x(4) x(2) x(6) x(1) x(5) x(3) x(7) X(0) X(1) X(2) X(3) X(4) X(5) X(6) X(7) -1 -1 -1 -1 -1 -1 J J -J -J -1 -1 -1 -1 W W W 2 1 DIT/SRFFT This graph looks like the transpose of an N-point DIF FFT. The twiddle factors

come before the second stage. 3 Figure 8.15-1: 8.16 x = xR + jxI = (a + jb)(c + jd) e = (a b)d 1 add , 1 mult xR = e + (c d)a 2 adds 1 mult xI = e + (c + d)b 2 adds 1 mult Total 5 adds 3 mult 8.17 N-1 X(z) = . x(n)z -n n=0 246 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N-1 2 0,1,...,N 1 are the N sample points. It is clear that X(zk),k = N -n -j kn Hence, X(zk)= x(n)r e n=0 2 -n 0,1,...,N 1is equivalenttotheDFT(N-pt) of thesequence x(n)r,n . [0,N 1]. N -j k,k = where zk = re 8.18 x ' (n)= = = = Therefore Lx ' (Ln)= = LN-11 . ' (k)W-kn X LN LN

k=0 "k0-1 1 . LN k=0 "k0-1 1 . LN k=0 "k0-1 1 . LN k=0 "k0-1 1 . LN-1 ' (k)W-kn ' (k)W-kn X + X LN X(k)W-kn + LN X(k)W-kn + LN X(k)W-kn + N N k=0 k=N-k0+1 x(n)

LN k=LN-k0+1 LN-1 X(k+ N LN)W-kn LN k=LN-k0+1 N-1 . -(k-N+LN)n X(k)W LN k=N-k0-1 N-1 . X(k)W-kn N L =1 is a trivial case with no zeros inserted and 11 11 1. x ' (n)= x(n)= , + j ,0, j 22 22 2 8.19 N-1 x(n)Wkn X(k)= N n=0 Let F(t),t =0,1,...,N 1 be the DFT of the sequence on kX(k). F(t) = N-1

. k=0 = N-1 . k=0 = N-1 . n=0 = N-1 . n=0 X(k)Wtk N "N-1 . x(n)Wkn Wtk NN n=0 "N-1 . . k(n+t) x(n) W N k=0 x(n)d(n + t)mod N 247 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

N-1 = x(n)d(N 1n t) t =0,1,...,N 1 n=0 = {x(N 1),x(N 2),...,x(1),x(0)} 8.20 2N-1 y(n)Wkn Y(k)= k =0,1,..., 2N 1 N n=0 2N-1 y(n)Wkn = 2N n=0,n even N-1 y(2m)Wkm = N m=0 N-1 x(m)Wkm = N m=0 = X(k),k . [0,N 1] = X(kN),k . [N,2N 1]

8.21 (a) 12pn w(n)= (1cos ), 0 = n = N 1 2N 111 j -j = (eN 2n -1 + e N 2n -1 ) 24 N-1 -n W(z)= w(n)z n=0 N-1 . 1 1 j -j -n = (eN 2n -1 + e N

2n -1 ) z 24 n=0 -N -1 11z11(ze j 2 N-1 )N = 21z-1 41z-1e-1 j 2 N-1 11(ze -j 2 N-1 )N 41z-1e -j 2 N-1 (b) xw(n)= w(n)x(n) . Xw(k)= W(k)NX(k) 8.22 The standard DFT table stores N complex values WN k ,k =0,1,...,N 1. However, since NN k+ k+ 2 N 4 W = -Wk , we need only store Wk k =0,1,..., 1. Also, W = -jWk which is

NNN 2 NN 248 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

merely an interchange of real and imaginary parts of Wk and a sign reversal. Hen ce all essential N N quantities are easily obtained from Wk k =0,1,..., 1 N 4 8.23 The radix-2 FFT algorithm for computing a 2N-pt DFT requires 2N log22N = N + Nlo g2N N NN complex multiplications. The algorithmin(8.2.12)requires2[N log2N+ ]= +log2N complex 2 22 multiplications. 8.24 PM -k k=0 bkz since H(z)= 1+PN k=1 akz-k PM bkWkn 2p k=0 N+1 H( k)= N 1 1+PN k=1 akWkn N+1

. = H(k),k =0,...,N Compute N +1-pt DFTs of sequences {b0,b1,...,bM,0,0,..., 0} and {1,a1,...,aN} (assumes N >M), sayB(k)and A(k) k =0,...,N B(k) H(k)= A(k) 8.25 8 y(n)Wnk 9 Y(k)= n=0 y(n)Wnk 9 y(n)Wnk 9 + y(n)Wnk 9 + = n=0,3,6 n=1,4,7 n=2,5,8 22 2 y(3m)W3km 9 + y(3m +1)W (3m+1)k 9

+ (3m+2)k 9 y(3m +2)W = m=0 m=0 m=0 22 2 y(3m)Wkm 3 + y(3m +1)W3 mk W9 k + W2k y(3m +2)Wmk 39 = m=0 m=0 m=0 Total number of complex multiplies is 28 and the operations can be performed inplace. see fig 8.25-1 8.26 8 x(n)Wnk 9 X(k)= n=0 249 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(1) X(3) x(2) X(6) x(0) x(3) x(6) X(0) X(1) X(2) W 3 1 W 3 2 W 3 W 3 W 3 1 W 3 2 W 3 W 9 1 W 9 2 W 9 2 W 9 1 1 2 1 x(4) X(4) x(5) X(7) x(7) X(5) x(8) X(8) Figure 8.25-1: 58

2 . x(n)Wkn 9 + x(n)Wnk 9 + x(n)Wnk 9 = n=0 n=3 n=6 22 2 x(n)Wkn 9 + x(n +3)W9 nk W3 k + W2k x(n +6)Wnk 93 = n=0 n=0 n=0 22 2 x(n)Wnl 3 + x(n +3)Wnl 3 + x(n +6)Wnl

3 x(3l) = n=0 n=0 n=0 22 2 x(n)Wnl 3 W n 9 + x(n +3)W3 nl W9 nW31 + n x(n +6)W3 nl W9 W2 3 x(3l+1) = n=0 n=0 n=0 2 x(n)+W31 x(n +3)+ W32 x(n +6) Wnl 3 Wn 9 = n=0 2 x(n)+W32 x(n +3)+ W31 x(n +6) Wnl 3 W2n 9

x(3l+2) = n=0 Thenumberof required complexmultiplicationsis28. Theoperations canbeperformedinplace. see fig 8.26-1 8.27 (a)Refer to fig 8.27-1 (b)Refer to fig 8.27-2 250 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(0) W3 1 1 W 3 2 W 3 2 W 3 1 W 3 1 W3 2 W9 W3 2 W3 W 9 11 W 3 2 W3 2 2 W3 W 3 2 W 3 1 1 W9 W3 1 2 W9 4 W 3 1 W3 2 W3 1 W3 1 2 X(0) x(1) X(3) x(2) X(6)

x(3) X(1) x(4) X(4) x(5) X(7) x(6) X(2) x(7) X(5) x(8) X(8) W W33 Figure 8.26-1: (c)DIFispreferablefor computingallpoints. Itisalsobetterwhenonly X(0),X(1),X(2), X(3) are to be calculated. The rule is to compare the number of nontrivial complex mu ltiplies and choose the algorithm with the fewer. (d) IfM <<N and L <<N , the percentage of savings is N log2N-ML log2N 22 100% =(1ML ) 100% N log2NN 2 251 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(0) 0W16 0W16 0W16 0W16 0W16 0W16 0W16 0W16 0W16 0W16 0W16 0W16 W16 4 W16 4 W16 4 W16 4 -1 -1 -1 -1 -1 -1 -1 X(0) x(1) X(8) X(4) X(12) X(2) X(10) X(6) X(14) X(1) X(9) X(5) X(13) X(3) X(11) X(7) X(15) -1 Figure 8.27-1: x(0) X(0) X(1)

X(2) -1 -1 -1 -1 -1 -1 -1 W16 W16 W16 W16 W16 W16 W16 W16 0 1 2 3 4 5 6 7 X(3) X(4) X(5) X(6) X(7) x(1) X(8) X(9) X(10) X(11) X(12) X(13) X(14) X(15)-1 Figure 8.27-2: 252 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8.28 (a)Refer to fig 8.28-1. If data shuffling is not allowed, then X(0),...,X(3)shou ld be computed x(0) W16 W16 4 W16 W16 0 W16 0 W16 4 W16 0 W16 W16 W16 4 0 2 0 W16 0 W16 0 6 -1 -1 -1 -1 WWW16W -1 16 16 16 W WW1616 4 16 WW1616 16 W W16 0 WW16W 1616 4 16 0 4 0 2 0 0 W16160 67 -1 -1 -1 1616-1 W0 16W0

W W W 1 2 3 4 5 6 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 X(0) x(1) X(8) x(2) X(4) x(3) X(12) x(4) X(2) x(5) X(10) x(6) X(6) x(7) X(14) x(8)

X(1) X(9) x(9) x(10) X(5) x(11) X(13) x(12) X(3) x(13) X(11) x(14) X(7) x(15) -1 X(15) -1-1 -1 Figure 8.28-1: by one DSP. Similarly for X(4),...,X(7)and X(8),...,X(11)and X(12),...,X(15). Fr om the flow diagram the output of every DSP requires all 16 inputs which must therefore be stored in each DSP. (b)Refer to fig 8.28-2 (c)The computations necessaryforageneralFFT areshowninthefigureforpart(a), Ng = N log2N. Parallel computation of the DFTs requires 2 p-1 1 NN . N 1 Np = log2 + 2MM 22i

i=1 NN 1 = log2 + N(1) 2MM M Complexoperations,asisseen inthe figurefor(b). Thus Ng S = Np N log2N 2 = NN 1 log2 + N(1) 2MM M Mlog2N = log2N log2M +2(M 1) 253 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(0) x(1) x(2) x(3) x(4) W16 W16 WWW4 W0 16 W16 4 16 16 16 W0 16 0 2 6 -1 -1 -1 -1 0 -1 -1 -1 -1 X(2) x(5) X(10) x(6) X(6) x(7) X(14) x(8) x(9) x(10) x(11) x(12) x(13) x(14) x(15) Figure 8.28-2: 8.29 Refer to fig 8.29-1 N-1

1 . X(k)W-kn x(n)= N N k=0 1 . X(k)W-kn = 8 8 k even 3 1 . X(2m)W-mn = 4 8 m=0 3 1 . . X(k)W-kn 8 k odd 3 . W-n X(2m +1)W-mn 48 = X(2m)+X(2m +1)W-n 8

8 m=0 . 3 1 . x(n)= 8 m=0 . 3 1 . x(n +4) = 8 m=0 1 + 8 1 + 8 m=0 X(2m)W-mn 4 X(2m)W-mn 4 3 + W-n . 8 m=0 3 W-n .

8 m=0 W-mn 4 X(2m +1)W-mn , 0 = n = 3 4 X(2m +1)W-mn , 0 = n = 3 4 Thisresult canbeobtainedfromtheforwardDITFFT algorithmby conjugating each occurr ence 1 of Wi . W-i and multiplying each output by 1 (or can be multiplied into the outputs of NN 82 each stage). 254 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 -1 W8 W8 -0 -2 -1 -1 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -1 -1 W8 W8 -0 -2 -1 -1 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -1 -1 -1 -1 W8 -0 W8 W8 W8 -1 -2 -3 X(0) X(4) X(2) X(6) X(1) X(5) X(3) X(7) x(0) x(1) x(2) x(3) x(4) x(5) x(6)

x(7) -1 -1 W8 W8 -0 -2 -1 -1 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -1 -1 W8 W8 -0 -2 -1 -1 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -1 -1 -1 -1 W8 -0 W8 W8 W8 -1 -2 -3 X(0) X(4) X(2) X(6) X(1) X(5) X(3) X(7) x(0) x(1) x(2) x(3) x(4) x(5) x(6)

x(7) Figure 8.29-1: 8.30 7 1 X(k)W-kn 8 x(n)= 8 k=0 37 1 8 X(k)W-kn 1 + 8 8 X(k)W-kn 8 = k=0 k=4 . 33 1 X(k)W-kn 8 +(-1)n X(k+4)W-kn 8 = 8

k=0 k=0 . 33 1 X(k)W-lk 4 + X(k+4)W-lk 4 x(2l) l =0,1,2,3 = , 8 k=0 k=0 . 33 1 X(k)W-lk W-k 48 W-k X(k+4)W-lk 48 x(2l+1) , l =0,1,2,3 = 8 k=0 k=0 SimilartotheDIT case(prob.8.29) result canbeobtainedby conjugating each Wi and s

caling N by 18. Refer to fig 8.30-1 255 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-1 -1 -1 -1 W8 W8 -0 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -2 -1 -1 W8 W8 -0 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -2 -1 -1 -1 W8 -0 W8 W8 W8 -1 -2 -3 X(2) X(3) X(4) X(5) X(6) X(7) X(0) X(1) x(0) x(4) x(2) x(6) x(1) x(5) x(3) x(7) -1 -1

-1 -1 W8 W8 -0 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -2 -1 -1 W8 W8 -0 W8 -0 W8 -0 1/8 1/8 1/8 1/8 -2 -1 -1 -1 W8 -0 W8 W8 W8 -1 -2 -3 X(2) X(3) X(4) X(5) X(6) X(7) X(0) X(1) x(0) x(4) x(2) x(6) x(1) x(5) x(3) x(7) -1 -1-1 Figure 8.30-1: 8.31

x(n)= x * (N n) N-1 1 . * (n)W-kn IDFT(x * (n)) = x N N n=0 N-1 1 . x(N n)W-kn = N N n=0 1 1 . -k(N-m) = x(m)W N N m=N N-1 1 . ' )W-km ' = x(N m N N m ' =0 Since the IDFT of a Hermitian symmetric sequence is real, we may conjugate all t erms in the

sum yielding N-1 IDFT(x * (n)) = 1 . x * (N m ' )Wkm ' N N m ' =0 N-11 . x(n)Wkn = N N n=0 256 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 = X(k) N Ingeneral,theIDFT of anN-length sequence canbeobtainedby reversingthe flowof afo rward 1 FFT andintroducing a scalefactor . Since theIDFTisapparently capableofproducingthe N (scaled)DFTforaHermitiansymmetricsequence,thereversed flowFFTwillproducethe desi red FFT. 8.32 N-1 x(m)Wkm X(k)= N m=0 N-1 x(m)Wkm W-kN = . since W-kN =1 NN N m=0 N-1 . -k(N-m) = x(m)W N m=0 This can be viewed as the convolution of the N-length sequence x(n)with the impu sle response of a linear filter. hk(n) . = Wkn N u(n), evaluated at time N 8 Hk(z) = .

Wkn N z -n n=0 1 = 1Wk Nz-1 = Yu(z) X(z) yk(n) = Wk Nyk(n 1)+x(n), yk(-1) = 0 yk(N) = X(k) 8.33 (a)11frequencypointsmustbe calculated.Radix-2FFT requires 1024 log21024 5000 complex 2 multiplies or 20,000 real multiplies. FFT of radix-4 requires 0.75 5000 = 3,750 complex multiplies or 15,000 real multiplies. Choose Goertzel. (b) In this case, direct evaluation requires 106 complex multiplies, chirp-z 22 103 comples multiplies, andFFT1000+ 5000 13 =33 103 complex multiplies. Choose chirp-z. 2 8.34 N In the DIF case, the number of butterflies affecting a given output is in the first stage, N in 24 the second, .... The total number is 1+2+... +2-1 =2-(1( 1))= N 1 2 257 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Everybutterflyrequires4 real multiplies,andthe erorvarianceis 2 Under the assumption that 12. the errors are uncorrelated, the variance of the total output quantization error is d2 Nd2 sq 2 =4(N 1) = 12 3 8.35 (a) 11 X * Re[Xn+1(k)] = Xn+1(k)+ n+1(k) 22 11 11 W-mX * Wm X * = (k)+ (l)+ (k)(l) XnN XnnNn 22 22 m = Re[Xn(k)]+Re[WN Xn(l)] 11

since |Xn(k)| < , |Re[Xn(k)]| < 22 11 m since |Xl(k)| < , |Re[WN Xn(l)]| < 22 1 m so |Re[WN Xn(l)]| < 2 m Therefore |Re[Xn+1(k)]|=|Re[Xn(k)]| + |Re[WN Xn(l)]| < 1 The other inequalities are verified similarly. (b) Xn+1(k) = Re[Xn(k)]+jIm[Xn(k)] 2p 2p [cos( m)jsin( m)][Re[Xn(l)]+jIm[Xn(l)]] NN = Re[Xn(k)]+cos(.)Re[Xn(l)]+sin(.)Im[Xn(l)] +j{Im[Xn(k)]+cos(.)Im[Xn(l)]+sin(.)Re[Xn(l)]} Therefore, |Xn+1(k)| = |Xn(k)| + |Xn(l)| + A . where A =2cos(.){Re[Xn(k)]Re[Xn(l)]+Im[Xn(k)]Im[Xn(l)]} +2sin(.){Re[Xn(k)]Im[Xn(l)]Im[Xn(k)]Re[Xn(l)]}

also |Xn+1(l)|2 = |Xn(k)|2 + |Xn(l)|2 A(*) Therefore, if A = 0, max[|Xn+1(k)|,|Xn+1(l)|]= |Xn+1(k)| 1 . 2 = |Xn(k)|2 + |Xn(l)|2 + A> max[|Xn(k)|,|Xn(l)|] By similarmeans using(*),it canbeshownthatthesameinequality holdsif A< 0. Also, from the pair of equations fro computing the butterfly outputs, we have 2Xn(k)= Xn+1(k)+Xn+1(l) W-m 2Xn(l)= Xn+1(k)W-mXn+1(l) NN By a similar method to that employed above, it can be shown that 2max[|Xn(k)|,|Xn(l)|]= max[|Xn+1(k)|,|Xn+1(l)|] 258 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) N=64 dc=16 (b) N=64 dc=8 20 8 15 6 10 4 magnitude magnitude magnitude 5 2 0 0 0 20406080 0 20406080 (c) N=128 dc=16 (d) N=64 dc=7.664e-14 20 800 15 600 magnitude 10 400 5 200 0 0 0 50 100 150 0 20406080 Figure 8.36-1: 8.36

Refer to fig 8.36-1. (d)(1)Thefrequencyintervalbetweensuccessivesamplesfortheplotsin parts(a),(b),(c) and 11 (d) are 1 , , and 1 respectively. 646412864 (2)The dc values computed theoretically and from the plots are given below: part a part b part c part d theoretical 16 8 16 0 practical 16 8 16 8.203e 14 Both theoretical andpracticaldcvaluesmatch exceptinthelast casebecauseof the fin iteword length effects the dc value is not a perfect zero.

(3)Frequency interval = N1 . (4)Resolution is better with N =128. 8.37 (a)Refer to fig 8.37-1. (b) (c) (d) (e) 259 Refer Refer Refer Refer to to to to fig fig fig fig 8.37-1. 8.37-1. 8.37-1. 8.37-2.

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

r=0.9, Y(k) r=0.9, c=0.92, W(k) r=0.9, c=0.92, W(k) magnitude 0 1 2 3 4 5 6 magnitude 0 5 10 15 20 25 magnitude 0 50 100 150 0 50 100 150 r=0.5, Y(k) r=0.5 , c=0.55, W(k) 0.8 1 1.2 1.4 magnitude 0 1 2 3 4 5 6 magnitude 0 50 100 150 0 50 100 150 Figure 8.37-1: r = 0.5, Y(k) 32 W(k) x 10 12 4 10 3 8 magnitude 6

2 4 1 2 0 0 0 50 100 150 0 50 100 150 Figure 8.37-2: 260 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 9 -5 -6 (a)H(z)=1+2z-1 +3z-2 +4z-3 +3z-4 +2z+ z. Refer to fig 9.1-1 -4 -5 (b)H(z)=1+2z-1 +3z-2 +3z-3 +2z+ z. Refer to fig 9.1-2 x(n) z-1 z-1 z-1 z-1 z-1 z-1 + + + 3 2 4 y(n) Figure 9.1-1: 9.2 Refer to fig 9.2-1 A4(z)= H(z) B4(z) Hence, K4 A3(z) = = = = = 1+ 2.88z -1 + 3.4048z -2 + 1.74z -3 + 0.4z -4 0.4+ 1.74z -1 + 3.4048z -2 + 2.88z -3 + z -4 0.4 A4(z)k4B4(z) 1k2 4 1+ 2.6z -1 + 2.432z -2 + 0.7z -3 261 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ x(n) z-1 z-1 z-1 z-1 -1z+ + 2 3 + x(n) z-1 z-1 z-1 z-1 -1z+ + 2 3 y(n) Figure 9.1-2: -2 -3 B3(z) =0.7+2.432z -1 +2.6z + z Hence, K3 =0.7 A3(z)k3B3(z) A2(z)= 1k32 -2 = 1+1.76z -1 +1.2z -1 -2 B2(z) =1.2+1.76z + z Then, K2 =1.2 A2(z)k2B2(z) A1(z)= 1k22 -1 = 1+0.8z Therefore, K1 =0.8 Since K2 > 1, the system is not minimum phase. 9.3 V(z)= X(z)+ 1 z -1V(z) 2 1 v(n)= x(n)+ v(n 1) 2

Y(z) = 2[3X(z)+V(z)]+2z -1V(z) Y(z) H(z)= X(z) -1 8z = 10.5z-1 h(n) = 8(0.5)n u(n)(0.5)n-1 u(n 1) 262 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ z-1 z-1 z-1 z-1 + + + y(n) 1 2.88 3.4048 1.74 0.4 x(n) + z-1 z-1 z-1 z-1 + + + y(n) 1 2.88 3.4048 1.74 0.4 x(n) (a) z-1 + z-1 ++ z-1 + ++ z-1 + + x(n) k1 k2 k3 k4 f1(n) f 2(n) f 3(n) f 4 (n) = y(n) (b) Figure 9.2-1: (a) Direct form. (b) Lattice form 9.4 3z1 1+2z1 H(z) =5+ + 1z-1 11z-1 1+ 32 1 11 )n-1 )n-1 h(n) =5d(n)+3(u(n 1)+( )n u(n)+2( u(n 1) 322 9.5

9z1 5 -2 6+ z 23 H(z)= 1 (1+ z-1)(11z-1) 329z1 5z-2 6+ 23 = 11z1 1z-2 66 Refer to fig 9.5-1 9.6 11 For the first system, H(z)= + 1b1z-1 1b2z-1 -1 1(b1 + b2)z H(z)= (1b1z-1)(1b2z-1) c0 + c1z-1 For the second system, H(z)= (1-

d1z-1)(1a2z-1) clearly, c0 =1 263 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

z-1 + z-1 + + 6 9/2 -5/3 1/6 1/6 y(n)x(n) z-1 + z-1 + + 6 9/2 -5/3 1/6 1/6 y(n)x(n) Figure 9.5-1: c1 = -(b1 + b2) d1 = b1 a2 = b2 9.7 (a) y(n)= a1y(n 1)+a2y(n 2)+b0x(n)+b1x(n 1)+b2x(n 2) -1 -2 b0 + b1z+ b2z H(z)= 1+a1z-1 + a2z-2 (b) -1 -2 1+2z+ z H(z)= 1+1.5z-1 +0.9z-2 Zeros at z = -1,-1 Poles at z = -0.75 j0.58 Since the poles are inside the unit circle, the system is stable. -1 -2

1+2z+ z H(z)= 1+z-1 2z-2 Zeros at z = -1,-1 Poles at z =2,-1 264 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The system is unstable. (c) p x(n)= cos( n) 3 1 H(z)= 1+z-1 0.99z-2 1 H(w)= 1+e-jw 0.99e-j2w p -j H( ) = 100e 3 3 pp Hence, y(n) = 100cos( n ) 33 1 y(n)= y(n 2)+x(n) 4 1 H(z)= 1-

1 z-2 4(a) h(n) = 1 2 . ( 1 2)n + (1 2)n . u(n) 1 1 H(z) = 2 11 2z-1 + 2 1+ 1 2z-1 (b) x(n) = . ( 1 2)n + (1 2)n . u(n) 1 1 X(z) = 11 2z-1 + 1+ 1 2z-1 2 X(z) = 11 4z-2 Y(z) = X(z)H(z) = 1 1+ 1 2z-1 + 1 11 2z-1 + 1 2z-1 (11 2z-1)2 + 1 2z-1

(1+1 2z-1)2 y(n) = . ( 1 2)n + (1 2)n n( 1 2)n + n(1 2)n . u(n) (c)Refer to fig 9.8-1 (d) 1 H(w)= 11e-j2w 4 4 sin2w -1 = v tan 178cos2w 4cos2w Refer to fig 9.8-2. 265 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form 2 cascade form x(n) y(n) x(n) + D + D 1/2 -1/2 y(n) + D D 1/4 Parallel form + + D + D 1/2 1/2 -1/2 y(n) x(n) 1/2 Figure 9.8-1: Magnitude of H(w) 0 0.5 1 1.5 2 2.5 3 3.5 ---> w Phase of H(w) 0.8 1 1.2 1.4 ---> |H(w)| ---> angle of H(w) 0.2 0 -0.2 0 0.5 1 1.5 2 2.5 3 3.5 ---> w Figure 9.8-2: 266

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(a) 1 -1 1+ z 3 H(z)= 1 z-2 13 z-1 + 481 -1 1+ z 3 = (11 z-1)(11 z-1) 2410 7 33 = 11 z-1 + 11 z-1 24 Refer to fig 9.9-1 (b) + z-1 + + z-1 z-1 + + z-1

z-1 + 1/3 y(n) x(n) 3/4 -1/8 3/4 -1/8 y(n) 1/3 x(n) Cascade: Parallel: + + + z-1z-1 + z-1 z-1 + + y(n) 1/2 1/3 1/4 x(n) 1/4 1/2 10/3 -7/3 x(n) y(n) Figure 9.9-1: H(z) = = 0.7(10.36z-2) 1+ 0.1z-1 0.72z-2 0.7(10.6z-1)(1+ 0.6z-1) (1+ 0.9z-1)(10.8z-1) = 0.350.1647 1+ 0.9z-1 0.1853 10.8z-1 Refer to fig 9.9-2 (c) H(z) = 3(1+ 1.2z-1 + 0.2z-2) 1+ 0.1z-1 0.2z-2 267 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1. Direct form I: Direct form II:

Direct form II: Direct form I: + + + z-1 z-1 + + z-1 z-1 + z-1 z-1 y(n) x(n) y(n) -0.36 -0.1 0.72 -0.1 0.72 -0.36 0.7 x(n) Cascade: Parallel: + + + z-1z-1 + z-1 z-1 + + +x(n) x(n) y(n)-0.9 -0.6 0.8 0.6 0.7 y(n) 0.8 -0.9 0.35 -0.1853 0.1647 Figure 9.9-2: 3(1+0.2z-1)(1+z-1) = (1+0.5z-1)(10.4z-1) 71 = -3+ 10.4z-1 1+0.5z-1

Refer to fig 9.9-3 (d) v -1 -2) 2(1z-1)(1+ 2z+ z H(z)= (1+0.5z-1)(10.9z-1 +0.8z-2) vv 2+(2 22)z-1 +(2 2 2)z-2 2z-3) = 10.4z-1 +0.36z-2 +0.405z-3 AB + Cz-1 =+ 1+0.5z-1 10.9z-1 +0.8z-1 Refer to fig 9.9-4 (e) -1 1+z H(z)= 11 z-1 1 z-2 24-1 1+z = (10.81z-1)(1+0.31z-1) 1.62 -0.62 =+

10.81z-1 1+0.31z-1 Refer to fig 9.9-5 -1-2 1-z +z (f) H(z)= . Complex valued poles and zeros.Refer to fig 9.9-6 All the above 1-z-1+0.5z-2 268 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form II: Direct form I: + + z-1 z-1 + + -0.1 0.2 0.2 1.2 3 y(n) x(n) y(n) + + + z-1 z-1 z-1 z-1 + -0.1 x(n) 1.2 0.2 0.2 3 Cascade: Parallel: x(n) + + + z-1z-1 + -0.5 0.2 0.4 1 3 + z-1 z-1 + + x(n) y(n) 7 -3 -1 -0.5 0.4 y(n) Figure 9.9-3:

systems are stable. 269 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form II: Direct form I: 3 + + z-1 z-1 + + + + z-1 2 -1 1-2 0.4 -0.36 -0.405 y(n) -1 + + + z-1 z-1 z-1 z-1 + + + z-1 2 2 x(n) 3 0.4 -0.36 -0.405 1-1 Parallel: + + + z-1 + z-1 z-1 + z-1 + x(n) Cascade: y(n) -0.5 y(n) -1 x(n) 2 -1 0.9 1.414

y(n) -0.81 1 + z-1 z-1 + + z-1 + + x(n) -3 -0.8 0.9 C B A -0.5 Figure 9.9-4: Direct form II: Direct form I: + z-1 + + z-1 z-1 + + z-1 z-1 +y(n) x(n) y(n) 1/2 1/4 1/2 1/4 1 x(n) Cascade: Parallel: + + + z-1z-1 + z-1 z-1 + + y(n)x(n) x(n) y(n)0.81 1 -0.31 0.81 -0.3 -0.62

1.62 Figure 9.9-5: 270 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form I: Direct form II, cascade, parallel: + + y(n) x(n) + + + z-1 z-1 z-1 z-1 +-1 1 1 y(n)x(n) z-1 z-1 + + 1 -1 -1/2 -0.5 1 Figure 9.9-6: 271 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.10 Refer to fig 9.10-1 + + + x(n) v(n) w(n) w(n-1) r sin w0 y(n) z-1 Figure 9.10-1: 1 H(z)= r2z-2 12rcosw0z-1 + (1) V(z)= X(z)rsinw0z -1Y(z) (2) W(z)= V(z)rcosw0z -1W(z) (3) Y(z)= rcosw0z -1Y(z)rsinw0z -1W(z) By combining(1) and(2) we obtain 1 rsinw0z-1 (4) W(z)= X(z)Y(z) 1rcosw0z-11rcosw0z-1 Use(4) to eliminate W(z)in(3). Thus, -1)2 -2] Y(z)[(1rcosw0z + r 2sin2 w0z = X(z) 22 Y(z)[12rcosw0z -1 +(r cos w0 + r 2sin2 w0)z -2]= X(z) Y(z)1 = r2z-2 X(z)12rcosw0z-1 + 9.11 A0(z)= B0(z) =1 -1 A1(z)= A0(z)+k1B0(z)z

1 -1 = 1+ z 2 272 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

B1(z) A2(z) B2(z) A3(z) B3(z) A4(z) Therefore, H(z) where C is a constant 9.12 Refer to fig 9.12-1 x (n) k x(n) = x1(n) Hk(z)= yk(n)= z-1 + z-1 + + b0k b1k b2k w2k(n) w1k (n) H1 (z) H2(z) 1 -1 = + z 2 = A1(z)+k2B1(z) -2 = 1+0.3z -1 +0.6z -1 -2 =0.6+0.3z + z

= A2(z)+k3B2(z) -3 =10.12z -1 +0.39z -2 0.7z -2 -3 = -0.7+0.39z -1 0.12z + z = A3(z)+k4B3(z) 53 1 -3 -4 =1z -1 +0.52z -2 0.74z + z 150 3 53 1 -4) -3 = C(1z -1 +0.52z -2 0.74z + z 150 3 y(n) k -a 1k -a2k HN(z)xN(n) y(n)=y(n) N Figure 9.12-1: -1 -2

b0k + b1kz+ b2kz 1+a1kz-1 + a2kz-2 b0kxk(n)+w1k(n 1) 273 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

w1k(n)= b1kx(n)a1kyk(n)+w2k(n 1) w2k(n)= b2kx(n)a2kyk(n) 9.13 YJM1 =G *XIN DO 20 J=1,K YJ=B(J,0) * XIN + W1(J) W1(J) = B(J,1)*XIN -A(J,1)*YJ + W2(J) W2(J) = B(J,2)*XIN -A(J,2)*YJ YJM1 = YJM1 + YJ 20 CONTINUE YOUT = YJM1 RETURN 9.14 YJM1 = XIN DO 20 J=1,K W = -A(J,1) * WOLD1 -A(J,2) * WOLD2 + YMJ1 YJ = W + B(J,1)*WOLD1 + B(J,2)*WOLD2 WOLD2 = WOLD1 WOLD1 = W YJM1 = YJ 20 CONTINUE YOUT = YJ RETURN 9.15 1 -1 -2 H(z)= A2(z) = 1+2z + z 3 1 -1 -2 B2(z)= +2z + z

3 1 k2 = 3 A2(z)k2B2(z) A1(z)= 1k22 3 -1 = 1+ z 2 3 k1 = 2 9.16 (a) . . . . 1 -1 A1(z)1 k1 1 1+ z 2 == -11 -1 B1(z) k1 1 z + z 2 . . . . 1 -1 1 -2 A2(z)1 1 A1(z) 1+

z z 3 33 == 1 -1 -2 1 1 z-1B1(z) 1 + z+ z B2(z) 3 33 274 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. . A3(z) 11 A2(z) = B3(z) 11 z-1B2(z) -3 . H1(z)= A3(z)=1+z

j zeros at z = -1,e3 (b) H2(z)= A2(z)z -1B2(z) 22 -3 -1 -2 z = 1+ z z 33 v -5 j 11 The zeros are z =1, 6 (c) If the magnitude of the last coefficient|kN| = 1, i.e., kN = 1, all the zeros lie on the unit circle. (d)Referto fig9.16-1. Weobservethatthe filtersarelinear phase filterswith phasej umpsat 2 1

0 -1 -2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> freq(Hz) 4 2 0 -2 -4

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> freq(Hz) Figure 9.16-1: the zeros of H(z). 9.17 (a)Refer to fig 9.17-1 275 ---> phase of H2(w) ---> phase of H1(w) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

f (n) = y(n) 3 z-1 + z-1 + ++ z-1 + + x(n) f1(n) f 2(n) g 2 (n)g1 (n) 0.65 -0.34 0.8 g (n) 3 Figure 9.17-1: x(n)= d(n) f1(n)= d(n)+0.65d(n 1) g1(n) =0.65d(n)+d(n 1) f2(n)= f1(n)0.34g1(n 1) = d(n)+0.429d(n 1)0.34d(n 2) g2(n)= -0.34f1(n)+g1(n 1) = -0.34d(n)+0.429d(n 1)+d(n 2) h(n)= f3(n)= f2(n)+0.8g2(n 1) = d(n)+0.157d(n 1)+0.0032d(n 2)+0.8d(n 3) -3 (b)H(z)=1+0.157z-1 +0.0032z-2 +0.8z. Refer to fig 9.17-2 z-1 + + z-1 + z-1 0.157

0.0032 0.8 x(n) y(n) Figure 9.17-2: 276 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.18 (a) H(z) = A3(z) = B3(z) = k3 = A2(z) = = B2(z) = k2 = A1(z) = = B1(z) = k1 = C3(z) = D3(z) = k3 = C2(z) = = D2(z) = k2 = C1(z) = = D1(z) = k1 = C3(z) = = From the equations, we obtain v0 = v1 = v2 = v3 = C3(z) A3(z) -3 1+0.9z -1 0.8z -2 +0.5z -2 -3 0.50.8z -1 +0.9z + z 0.5 A3(z)k3B3(z) 1k32 -2 1+1.73z -1 1.67z -1 -2 -1.67+1.73z + z

-1.67 A2(z)k2B2(z) 1k22 -1 1+1.62z -1 1.62+z 1.62 -3 1+2z -1 +3z -2 +2z -2 -3 2+3z -1 +2z + z 2 C3(z)k3D3(z) 1k32 41 -1 -2 1+ z + z 33 14 -1 -2 + z + z 33 1 3 C2(z)k2D2(z) 1-

k22 3 -1 1+ z 4 3 -1 + z 4 3 4 v0 + v1D1(z)+v2D2(z)+ 1+2z 107 48 13 4 -1 2 -3 -1 +3z -2 +2z v3D3(z) 277 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The equivalent lattice-ladder structure is: Refer to fig 9.18-1 -3 (b)A3(z)=1+0.9z-1 0.8z-2 +0.5z, |k1| > 1 and |k2| > 1 . the system is unstable. x(n) + + z -1 + + z -1 + + z -1 + + + 0.5 -1.67 1.62 v3 = 2 v 2 =-1 v = -13/4 v 0 =-107/481 y(n) Figure 9.18-1: 9.19 Refer to fig 9.19-1 -1 Y(z) =[rsinX(z)+rcosY(z)rsinC(z)]z -1 C(z) =[-rcosX(z)+rsinY(z)+rcosC(z)]z Y(z) H(z)= X(z) -1 rsinz = r2z-2 Hence, h(n)= r nsin(n)u(n) 12rcosz-1 +

2 and y(n)= rsinx(n 1)+2rcosy(n 1)ry(n 2) j The system has a zero at z =0 and poles at z = re. 9.20 1 H(z)= r2z-2 rcos2w0 rcos2w0 12rcosw0z-1 + rcosw0 jrcosw0 + j 2sinw0 2sinw0 =1+ + z (rcosw0 + jrsinw0) z (rcosw0 jrsinw0) rcos2w0 rcosw0 j 2sinw0 S(z)= X(z) z (rcosw0 + jrsinw0) 278 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

r cos q + z-1 + z-1 x(n) -r cos q r sin y(n) r sin -r sin c(n) q q q r cos q Figure 9.19-1: s(n)= v1(n)+jv2(n) p = a1 + ja2 . a1 = rcosw0 a2 = rsinw0 A = q1 + jq2 . q1 = rcosw0 -rcosw0 q2 = 2sinw0 v1(n +1) = a1v1(n)a2v2(n)+q1x(n) = rcosw0v1(n)rsinw0v2(n)+rcosw0x(n) v2(n)= a2v1(n)+a1v2(n)+q2x(n) -rcosw0 = rsinw0v1(n)+rcosw0v2(n)+ x(n) 2sinw0 or, equivalently, rcosw0 -rsinw0 rcosw0 v(n +1) = v(n)+ rcosw0 x(n) rsinw0 rcosw0 2sinw0 y(n)= s(n)+s * (n)+x(n) =2v1(n)+x(n) or, equivalently, y(n) =[20]v(n)+x(n) where .

v1(n) . v(n)= v2(n) 279 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.21 (a) H(z)= (b) Refer to fig 9.21-1 Direct form: k1 = 0.6 A1(z) B1(z) A2(z) = = = = 1+ 0.6z -1 0.6+z -1 A1(z)+k2B1(z)z -1 1+ 0.78z -1 + 0.3z -2 B2(z) A3(z) = = = 0.3+ 0.78z -1 + z -2 A2(z)+k3B2(z)z -1 1+ 0.93z -1 + 0.69z -2 + 0.5z -3 B3(z) A4(z) = = = 0.5+ 0.69z -1 + 0.93z -2 + z -3 A3(z)+k4B3(z)z -1 1+ 1.38z -1 + 1.311z -2 + 1.337z -4 -3 +0.9z x(n) + z-1 zzz+ -1 -1 -1 + 1 1.38 1.311 1.337 0.9 y(n) Lattice form: y(n) z-1 + z-1 ++ z-1 + ++

z-1 + + x(n) 0.6 0.3 0.5 0.9 ---Figure 9.21-1: 9.22 (a) From (9.3.38) we have y(n) But, y(n) Hence, k2 and, k1(1+k2) k1 = = = = + -k1(1+k2)y(n 1)k2y(n 2)+x(n) 2rcosw0y(n 1)r 2 y(n 2)+x(n) r 2 -2rcosw0 2rcosw0 1+r2 280 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Refer to fig 9.22-1 (b) Whenr =1, the system becomes an oscillator. + + z -1 + + z -1 x(n) -y(n) k k 12 9.23 Figure 9.22-1: H(z) 1 For the all-pole system , we have A(z) k1(1+k2) k2 . k1 k2 For the all-zero system, C2(z) A2(z) B2(z) k2 A1(z) B1(z) k1 A0(z) 281 = = = = = = = = = = = =

= = = 10.8z-1 +0.15z-2 1+0.1z-1 0.72z-2 B(z) A(z) 0.1 -0.72 0.357 0.72 -2 10.8z -1 +0.15z -2 10.8z -1 +0.15z -1 -2 0.150.8z + z 0.15 A2(z)k2B2(z) 1k22 -1 10.696z -1 -0.696+z -0.696 B0(z)=1 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2 C2(z)= vmBm(z) m=0 = v0 + v1B1(z)+v2B2(z) -2 =10.8z -1 +0.15z The solution is: v2 =0.15 v1 0.18v2 = -0.8 v0 0.696v1 +0.15v2 =1 . v0 =1.5 v1 = -0.68 v2 =0.15 Thus the lattice-ladder structure is: Refer to fig 9.23-1 + + z -1 + + z -1 + + x(n) -0.15 -0.696 v 2 = 0.15 v 1 = -0.68 v 0 = 1.5 y(n) Figure 9.23-1: 9.24 v 2 -1-2 1z +0.25z H(z)= 2 Refer to fig 9.24-1 1-0.8z-1+0.64z-2 . 9.25

-1 1+z 1 H(z)= v 1-z-1 . 1-0.82z-1+0.64z-2 -1 2.31 -1.31+2.96z H(z)= 1-1 -1 + v Refer to fig 9.25-1 -2 2z1-0.82z-1+0.64z 282 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Direct form II : Transposed form : + D + + + D + D + D + a b cd y(n)x(n) where a=0.8, b = -2 /2, c = 0.25 and d = -0.64 b a dc x(n) y(n) Figure 9.24-1: 9.26 (a) For positive numbers, range is 21001 20111 01. 00... 0 01. 11... 1 |{. . |{. . 11 11 or 7.8125 10-3 2.5596875 102 negitive numbers 21001 20111 10. 11... 1 10. 00... 0 |{. . |{. .

11 11 or 7.8163 10-3 --2.56 102 (b) For positive numbers, range is 210000001 201111111 01. 00... 0 01. 11... 1 |{. . |{. . 23 23 or 5.8774717 10-39 3.4028234 1038 negitive numbers 210000001 201111111 10. 11... 1 10. 00... 0 |{. . |{. . 23 23 or 5.8774724 10-39 --3.4028236 1038 283 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Parallel: Cascade: + + + z-1 + z-1 z-1 x(n) 1/2 1 y(n) 0.8 2 -0.64 y(n) + z-1 z-1 + + z-1 + + x(n) 0.5 -0.64 0.8 2.96 -1.31 2.31 2 Figure 9.25-1: 9.27 (a)Refer to fig 9.27-1 x(n) y(n) -a 1 -a 2 + + -1 -1 z z Figure 9.27-1: -1 -2)-1 HR(z) = (1+a1z + a2z -a1 a2 4a2 1

poles zp1,2 = 2 for stability 2 (i)a1 4a2 = 0 2 -a1 a1 4a2 if a1 = 0, =-1 2 2 . a1 4a2 = 2a1 . a1 = 2 and a1 a2 = 1 284 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2 -a1 a1 4a2 if a1 < 0, = 1 2 . a2 4a2 = 2+a1 1 . a1 =-2 and a1 + a2 =-1 a1 4a2 a2 1 (ii)()2 +( ) = 1 22 a2 = 1 Refer to fig 9.27-2. The region of stability in the a1 a2 plane is shaded in the figure. There are The stable area of (a, a) 12 -1-2 1 2 a2 a 1 -1 2 1 Figure 9.27-2: nineinteger pairs(a1,a2)which satisfy the stability conditions. These are(with c orresponding system functions): -2)-1 (0,-1) HR1(z) =(1z (0,0) HR2(z) =1 -2)-1

(0,1) HR3(z) = (1+z -1)-1 (1,0) HR4(z) = (1+z -1 -2)-1 (1,1) HR5(z) = (1+z + z -1 -2)-1 (2,1) HR6(z) = (1+2z + z -1)-1 (-1,0) HR7(z) =(1z -1 -2)-1 (-1,1) HR8(z) =(1z + z -1 -2)-1 (-2,1) HR9(z) =(12z + z 285 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) HR1(z)= HR4(z)HR7(z) HR6(z)= HR4(z)HR4(z) HR9(z)= HR7(z)HR7(z) (c) Onlythe following cases can make h(n)FIR: (i) hR(n) = d(n) N Then H(z) = . z -i i=0 N y(n) = . x(n i) i=0 (ii) hR(n)* hF(n)= d(n) Then H(z) =1 y(n)= x(n) (d) see above. 9.28 Refer to fig 9.28-1 Note that4 multiplications and3 additions are required to implement H1(z). The a dvantage z-1 + z-1 + z-1 + x(n) b0 b 1 b 2 b 3 Figure 9.28-1: Structure of H1(z) of Horner s method is in evaluating H1(z)for a specific z0. Thus, if -1 -2 -3 H1(z)= b0 + b0b1z + b0b1b2z + b0b1b2b3z = b0 + z -1(b0b1 + z -1(b0b1b2 + z -1b0b1b2b3)) 286 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

the 3 multiplications and 3 additions are required for the evaluation of 9.1 in the field of z. If the various powers of z are prestored, then Horner s scheme has no advantage ov er the direct evaluation of 9.1. Refer to fig 9.28-2 This requires 4 multiplications and 3 additions. The linear-phase system is writ ten as +z-1z-1z-1 z-1 +z-1 + z-1 b 3 b 2 b 1 b0 -3 -2 -1 Figure 9.28-2: Structure of H(z)= b0z+ b0b1z+ b0b1b2z+ b0b1b2b3 2 -1 -2 -3 -4 H(z)= za3 + za2 + a1 + za0 + za1 + za2 + za3 By applying Horner s scheme, we can rewrite this as -1 H(z)= z 3(a3 + z -1(a2 + z -1(a1 + z -1(a0 + z -1(a1 + z -1(a2 + za3)))))) Assumingthatz-1 and z aregiven,adirect evaluationof H(z)at z = z0 requires8 mult iplications and 6 additions. Using Horner s scheme based on 9.28, requires the same number of operations as direct evaluation of H(z). Hence, Horner s scheme does not offer any savings in computation. 9.29 (a)When x1 and x2 are positive, the result is obvious. If x1 and x2 are negative , let x1 = -0n1 n2 ...nb = -0n1 n2 ... nb +000 ... 01 x2 = -0m1 m2 ...mb = -1m1 m2 ... mb +000 ... 01 x3 = x1 + x2 = -0n1 0 ... 0+0m1 0 ... 0+c where c = 00n2 ... nb +00 m2 ... mb +000 ... 010 If the sign changes, there are two possibilities (i) n1 = m1 =0 287 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. n1 = m1 =1 11 .|x1| > , |x2| > 22 .|x3| > 1, overflow (ii)(n1 =1,m1 =0,c =0) or (n1 =0,m1 =1,c =0) 1 1 . (|0n1 0 ... 0)10| > or (|0m1 0 ... 0)10| > 22 1 and |c10| > 2 .|x3| > 1, overflow (b) x1 = 0100 x2 = 0110 x3 = -0110 =1010 x1 + x2 = 1010, overflow x1 + x2 + x3 = 0100, correct result 9.30 (a) -1 -a + z H(z)= 1az-1 -jw -a + e |H(ejw)|2 = ||2

1ae-jw (-a + cosw)2 +(-sinw)2 = (1acosw)2 +(asinw)2 a2 2acosw +1 = =1 .w 12acosw + a2 (b) Refer to fig 9.30-1 (c) If|a | = |a |, where a means the quantized value of a, then the filter remains all-pass. x(n) y(n) z-1 a 1 -a+ + Figure 9.30-1: (d) Refer to fig 9.30-2 (e) Yes, it is stillall-pass. 288 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) y(n) a z-1 z-1 ++ + + Figure 9.30-2: 9.31 (a)y(n) = 2( 12 )n ( 14 )n. u(n) (b) Quantization table 1 x > 1x = 1 32 31 29 15 = x > x = 32 32 16 29 27 14 = x > x = 32 32 16 . . . 1 1 14 = x > x = 32 32 16 . . . 1 x < -1+ x = -1 32 . 4 1 . Therefore x(n) = 1

. , 16, 16,0,. . . , 0 8 y(n) = 16y(n 1)+x(n) . 12 7 3 1 . y(n) = 1 . , 16, 16, 16, 16,0,0,. . . (c) . 3 7 15 31 63 . y(n) = 1 . , 4, 16, 64, 256, 1024,. . . . 3 7 12 16 . y(n) = 1 . , 4, 16, 64, 256,0,0,. . . Errors occur when numberbecomes small. 289 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.32 y(n) =0.999y(n 1)+e(n) 1 1 e(n)is white noise, uniformly distributed in the interval 29 , Ey 2(n)= (10.9992)Ey 2(n)= = = Therefore, Ey 2(n)= = 9.33 29 0.9992Ey 2(n 1)+ Ee 2(n) Ee 2(n) . . 2 1 e 2de -. . 2 .2 2-8 where . = 12 11 1

)2 122810.9992 6.361x10-4 ( (a)poles zp1 =0.695,zp2 =0.180 Refer to fig 9.33-1 (b) Truncation + D + D x(n) y(n) 0.695 0.18 Figure 9.33-1: 0.695 0.180 poles zp1 =0.625,zp2 =0.125 (c) Rounding 0.695 0.180 poles zp1 =0.75,zp2 =0.125 290 5 . =0.625 8 1 . =0.125 8 6 . =0.75 8 1 . =0.125 8

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(d) |0.750.695| < |0.6950.625| Rounding is better |Ha(w)| = [(1.483+1.39cosw)(1.0324+0.36cosw)]|Hb(w)| = [(1.391+1.25cosw)(1.0156+0.25cosw)]1 2 1 2 |Hc(w)| = 9.34 (a) [(1.563+1.5cosw)(1.0156+0.25cosw)]1 2 H1(z) = 11 2 z -1 . 1. h1(n) = 1,2 H2(z) = (11 4 z -1)-1 h2(n) = ( 1 4)n u(n) H3(z) = (1+ 1 4 z -1)-1 h3(n) = (1 4)n u(n) Refer to fig 9.34-1 Cascade the three systems in six possible permutations to ob tain six realizaH (z) H1(z) H (z)

2 3 + z -1 + z -1 1/4 + z -1 -1/4-1/2 Figure 9.34-1: tions. 1 (b) Error sequence ei(n)is uniformly distributedoverinterval(12-b , 2-b). So s2 = 2-2b for 22ei 12 any i (call its2) e 291 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + + + + + z -1 z -1 z -1 1/4-1/2 -1/4 + + + + + + z -1 z -1 z -1 1/4-1/2 -1/4 ee 1(n) (n) e (n) 2 3 e e e e e e e e Figure 9.34-2: q q q q q (c) consider cascadeH1 -H2 -H3 Refer to fig 9.34-2 h4(n)= h2(n)* h3(n) 1 1)2 =1,0, ,0,( ,0,... 1616 88 s2 s2 h24(n)+ h23(n) s2 3.0745s2 3.0745s2

s2 3.3882s2 s2 3.2588s2 s2 3.2627s2 s2 3.3216s2 2 = n=0 n=0 21 + = 1 1(1 4 )2 )2 1( 16 q = using similar methods: s2 H1 H2 H3 = H2 H1 H3 = H2 H3 H1 =

H3 H1 H2 = H3 H2 H1 = 9.35 y(n)= Q[0.1d(n)]+Q[ay(n 1)] (a) y(n)= Q[0.1d(n)]+Q[0.5y(n 1)] 1 y(0) = Q[0.1] = 8 292 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) 9.36 (a)s2 x = rxx 1 y(1) = Q[ ]=0 16 y(2)= y(3)= y(4) =0 no limit cycle y(n) y(0) y(1) y(2) y(3)= y(4) limit cycle occurs 1 v (0)=3 . Ax = 3 = Q[0.1d(n)]+Q[0.75y(n 1)] 1 = Q[0.1] = 8 31 = Q[ ]= 328 31 = Q[ ]= 328 1

= 8 (b) 2-6 . = 2se .2 = 12 1 = 12 212 1 soSNR = 10log10 e e e q s2 = 10log10(12 212) = 46.91dB (c) left-justified. (d) 88 s2 s2 h2(n)+s2 h2(n) = 1 n=0 n=0 11 )2 e

Now s2 = ( 1 1228 . 1 16 and h2(n)= = 10.752 7 , n 16 . 11 11 . q q so s2 17 = 344,064 1 andSNR = 10log10 s2 )2 )2 ( + ( 12 26 = 28 7 12 = 43.06dB 293

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.37 .. Define .c = rcos.,.s = rsin. for convenience,(a) -.sy(n 1)+e1(n)+x(n)+.cv(n 1)+e2(n)= v(n) .sv(n 1)+e3(n)+.cy(n 1)+e4(n)= y(n) (b) -.sz -1Y(z)+E1(z)+X(z)+.cz -1V(z)+E2(z)= V(z) .sz -1V(z)+E3(z)+.cz -1Y(z)+E4(z)= Y(z) .sz-1 Y(z)= [X(z)+E1(z)+E2(z)] 12.cz-1 + r2z-21.cz-1 +[E3(z)+E4(z)] 12.cz-1 + r2z-2= H1(z)X(z)+H1(z)[E1(z)+E2(z)] +H2(z)[E3(z)+E4(z)] when H1(z)and H2(z)are as defined in the problem statement h1(n)= .s 1 r n-1sin.nu(n 1) sin. = r nsin(n.)u(n 1) = r nsin(n.)u(n) h2(n) = 1 r nsin(n +1).u(n)+.c 1 r n-1sin(.n)u(n 1) sin. sin. n r = d(n)+ [sin(n +1).cos.sin(n.)]u(n 1) sin.

n = d(n)+r cos(n.)u(n 1) n = r cos(n.)u(n) (c) .2 s2 s2 s2 s2 s2 ==== = ee1 e2 e3 e4 12 = 1(2-b)2 122-2b = 12 88 s2 2s2 = qe h21(n)+2s2 e h2 2(n) n=0 n=0 8 = 2s2 e 2n [r 2nsin2n. + r cos 2n.] = =

n=0 2s2 e 1 1r2n 2-2b 6 1 1r2n 9.38 (a) h1(n) = ( 1 2)n u(n) 294 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

h2(n) h(n) s2 q = = = = ( 1 4)n u(n) [2( 1 2)n ( 1 4)n]u(n) 2s2 e1 8 . n=0 h2 1(n)+ 2s2 e2 8 . n=0 h2 2(n) 64 35 s2 e1 + 16 15 s2 e2 (b) s2 q = = s2 e1 . n h2(n)+s2 e2 . n h2 1(n) 64 35 s2 e1 + 4 3 s2 e2 9.39 Refer to fig 9.39-1

z -1 z -1 + + x(n) y(n)1 a1 e (n) 1 aM-2 + a z -1 +M-1 e (n) M-2 e (n) M-1 Figure 9.39-1: 1 s2 ei = 12 2-2b .i s2 q = (M 1)s2 ei = (M 1) 2-2b 12 295 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.40 B(z) H(z)= A(z) . j (10.8e 4 )(10.8e -j 4 ) = G1 1 z-1) (10.5z-1)(1+ 3 -1) (1+0.25z-1)(15z 8 3 3 G2 j -j (1-

0.8e )(10.8e ) = H1(z)H2(z) (a) z -1 = e -jw At w = 0,z -1 = 1 =1 4 j 4 H1(w)|w=0 -j (10.8e )(10.8e 1 (10.5)(1+ ) 3 ) G1 =1 3 3 G1 =1.1381 (1+0.25)(15) 8

G2 =1 j -j )(10.8eG2 =1.7920 (b) Refer to (c) Refer to Refer to fig Refer to fig (10.8e ) 296 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1. fig 9.40-1. fig 9.40-2. 9.40-3. 9.40-4.

Direct for m I: + z -1 z -1 z -1 + z -1 + + -3/8 -5/32 x(n) 1 1/6 1/6 G y(n) Direct form II and cascade structure: x(n) z -1 + z -1 ++ + -3/8 -5/32 G 1/6 1/6 y(n) Figure 9.40-1: Direct form I, impulse response -1 -0.5 0 0.5 1 1.5 ---> mag 0 10 20 30 40 50 60 70 80 90100 Direct form I, step response 1.6 1.4 0 10 20 30 40 50 60 70 80 90100

1.2 1 0.8 ---> n Figure 9.40-2: 297 ---> mag (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-0.5 0 0.5 1 1.5 ---> mag ---> mag Direct form II, impulse response 0 10 20 30 40 50 60 70 80 90100 Direct form II, step response 1 1.1 1.2 1.3 1.4 1.5 ---> mag 0 10 20 30 40 50 60 70 80 90100 ---> n Figure 9.40-3: Cascade form , impulse response 1.5 1 0.5 0 0 10 20 30 40 50 60 70 80 90100 ---> mag -0.5 Cascade form, step response 1.5 1.4 0 10 20 30 40 50 60 70 80 90100 ---> mag 1.3 1.2 1.1 1 ---> n Figure 9.40-4: 298

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

9.41 (a) -3 8 k1 = 27 32 4 = 9 5 k2 = 32 Refer to fig 9.41-1a. (b) + z-1 + z-1 + z-1 f (n)1 k2k1 f (n) 0 + f 2(n) = y(n)x(n) (a) Forward + + z-1 + + z-1 k2 k1 f (n)1 f (n) 0 -y(n) (n) 0 g (n) f (n) 2 g2 x(n) Reverse (b)

Figure 9.41-1: 1 A(z)= 1 z-1) (10.5z-1)(1+ 3 1 = 11z-1 1z-2 661 k2 = 6 299 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 k1(1+k2)= 6 1 k1 = 5 Refer to fig 9.41-1b. (c) Refer to fig 9.41-2. (e) Refer to fig 9.41-3. + + z -1 + + z -1 + + x(n) -v 2 v 1 v 0 y(n) -1/6 -1/5 =-0.1563 =0.4336 =0.7829 Figure 9.41-2: (f)Finite word length effects are visible in h(n)for part f. 9.42 Refer to fig 9.42-1. 15 c = 16 9 H1(z) = 10 11 z-1 283 80 H2(z)= 1 z-1 1+

3 300 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

h(n) h(n) IR for part a IR for part b 1 1 0.5 0.5 0 0 -0.5 -0.5 0 50 100 0 50 100 IR for part c IR for part f 0.8 1 0.6 0.5 h(n) h(n) 0.4 0.2 0 0 -0.2 -0.5 0 50 100 0 50 100 --> n --> n Figure 9.41-3: 301 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Parallel form structure: y(n) c H1 (z) H(z) x(n) 2 Parallel form structure using 2nd-order coupled-form state-space sections + z -1 + z -1 + x(n) 1 1/2 A B -1/3 y(n) Figure 9.42-1: 302 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 10 10.1 (a)To obtain the desired length of 25, a delay of 25-1 =12 is incorporated into Hd(w). Hence, 2 p -j12w Hd(w) =1e, 0 =|w|= 6 =0, otherwise 1 6 hd(n)= Hd(w)e -jwndw 2p sin 6 6 (n 12) = p(n 12) Then, h(n)= hd(n)w(n) where w(n)is a rectangular window of length N =25. (b)H(w)= P24 h(n)e-jwn . plot |H(w)| and 6 H(w). Refer to fig 10.1-1. n=0

(c) Hamming window: np w(n) =(0.540.46cos ) 12h(n)= hd(n)w(n) for 0 = n = 24 Refer to fig 10.1-2. 303 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-150 -100 -50 0 50 ---> mag(dB)---> mag(dB) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) ---> phase 0 -2 -4 Figure 10.1-1: 50 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) -50 -100 -150 4 2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> phase 0 -2 -4 ---> Freq(Hz) Figure 10.1-2: 304 ---> mag(dB) (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(d) Bartlett window: 2(n 12) w(n) =10 = n = 24 24 Refer to fig 10.1-3. 0 -10

---> mag(dB) -20 -30 -40 -50 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) Figure 10.1-3: ---> phase 10.2 (a) p p -j12w Hd(w) =1e, |w|=

, =|w|= p 6 3 pp =0, =|w|= 63 . 1 -jwndw hd(n)= Hd(w)e 2p sin(n 12) sin(n 12) 36 = d(n)+ p(n 12) p(n 12) (b) Rectangular window: w(n) =1, 0 = n = 24 = 0, otherwise 305 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Refer to fig 10.2-1. (c) Hamming window: -50 -40 -30 -20 -10 0 10 ---> mag(dB) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 ---> phase 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 -2 -4 ---> Freq(Hz) Figure 10.2-1: np w(n) =(0.540.46cos ) 12 h(n)= hd(n)w(n) 24 -jwn H(w)= h(n)e n=0 Refer to fig 10.2-2. (d) Bartlett window: (n 12) w(n) =1, 0 = n = 24 12

Refer to fig 10.2-3. Notethatthe magnitude responsesin(c) and(d) arepoorbecausethetransition regionis wide. To obtain sharper cut-off, we must increase the length N of the filter. 306 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

---> mag(dB) ---> mag(dB) ---> phase ---> phase 5 0 -5 -10 -15 -20 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) Figure 10.2-2: 0 -5 -10 -15 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) Figure 10.2-3: 307 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.3 1 n (a)Hanning window: w(n)= (1cos ), 0 = n = 24. Refer to fig 10.3-1. 212 n n (b) Blackman window: w(n)=0.420.5cos +0.08cos . Refer to fig 10.3-2. 12 6 50 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> mag(dB) 0 -50 -100 -150 ---> Freq(Hz) 4 2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> phase 0 -2 -4 ---> Freq(Hz) Figure 10.3-1: 10.4 (a)Hanning window: Refer to fig 10.4-1. (b) Blackman window: Refer to fig 10.4-2.

The results are still relatively poor for these window functions. 10.5 p 1 M = 4, Hr(0)= 1, Hr( 2)= 2 Hr(w) = 2 M 2 -1 . h(n)cos[w( M 1 2 n)] n=0 = 2 1 . h(n)cos[w( 3 2 n)] n=0 308 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

---> mag(dB) 0 -50 -100 -150 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2

---> phase 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) Figure 10.3-2: 1 At w =0,Hr(0)=1 =2 h(n)cos[0] n=0 2[h(0)+h(1)] =1 (1) 1 pp 1 . p 3 At w = ,Hr( )= =2 h(n)cos[ ( n)]

222 22 n=0 -h(0)+h(1) =0.354 (2) Solving(1) and(2), weget h(0) =0.073 and h(1) =0.427 h(2) = h(1) h(3) = h(0) Hence, h(n)= {0.073,0.427,0.427,0.073} 309 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-15 -10 -5 0 5 ---> mag(dB)---> mag(dB) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) ---> phase ---> mag(dB) ---> phase 0 -2 -4 5 0 -5 Figure 10.4-1: 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) -10 -15 4 2 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 -2 -4 ---> Freq(Hz) Figure 10.4-2: 310 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.6 2pk M =15.Hr( )= 15 1,k =0,1,2,3 0,k =4,5,6,7 M 1 M-3 2 X M 1 Hr(w) = h( 2 )+ 2 h(n)cosw( 2 n) n=0 h(n) = h(M 1n) h(n) = h(14n) 6 Hr(w) Solving the above eqn yields, h(n) = = h(7)+ 2 n=0 h(n)cosw(7n) {0.3189,0.0341,-0.1079,-0.0365,0.0667,0.0412,-0.0498,0.4667 0.4667,-0.0498,0.0412,0.0667,-0.0365,-0.1079,0.0341,0.3189} 10.7 2pk .. . 1,k =0,1,2,3

)= 0.4,k =4 M =15.Hr( 15 0,k =5,6,7 M 1 M-3 2 X M 1 Hr(w) = h( 2 )+ 2 h(n)cosw( 2 n) n=0 h(n) = h(M 1n) h(n) = h(14n) 6 Hr(w)= h(7)+2 h(n)cosw(7n) n=0 Solving the above eqn yields, h(n) = {0.3133,-0.0181,-0.0914,0.0122,0.0400,-0.0019,-0.0141,0.52, 0.52,-0.0141,-0.0019,0.0400,0.0122,-0.0914,-0.0181,0.3133} 10.8 (a) ya(t) = dxa(t) dt = d dt[ej2Ft] = j2pFej2Ft Hence, H(F) = j2pF 311 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) |H(F)| 6 H(F) = = = 2pF p 2, p 2, F > 0 F < 0 Refer to fig 10.8-1. (c) B=pi/6 --> magnitude 0 0.2 0.4 0.6 -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 --> Freq(Hz) -2 -1 0 1 2 --> phase -0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 --> Freq(Hz) Figure 10.8-1: H(w)= jw, |w|= p |H(w)| = |w|p H(w)= , w> 0 2 p = , w< 0

2 Refer to fig 10.8-2. we note that the digital differentiator has a frequency response that resembles the response of the analog differentiator. (d) y(n)= x(n)x(n 1) -1 H(z) =1z -jw H(w) =1e 312 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

--> phase --> magnitude 4 3 2 1 0 -4-3-2-1 0 1 2 3 4 --> w 2 1 0 -1 -2 -4-3-2-1 0 1 2 3 4 --> w Figure 10.8-2: w w -j = e 2 (2jsin ) 2 w |H(w)| =2|sin | 2 pw H(w)= 22 Refer to fig 10.8-3.

w Note that for small w, sinw w and H(w) jwe-j2 , which is a suitable approximation 22 tothe differentiatorin(c). w0 (e) The valueH(w0)is obtainedfrom(d) above. Then y(n)= A|H(w0)|cos(w0n + .+ ) 22 10.9 -j10w Hd(w)= we , 0 = w = p -j10w = -we , p = w = 0 . 1 -jwndw hd(n)= Hd(w)e 2p cosp(n 10) = ,n 610 =

(n 10) =0,n =10 cosp(n 10) hd(n)= , 0 = n = 20,n 6=10 (n 10) 313 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.5 1 1.5 2 --> magnitude--> magnitude -4-3-2-1 0 1 2 3 4 --> w --> phase -2 -1 0 1 2 -4-3-2-1 0 1 2 3 4 --> w Figure 10.8-3: =0,n =10 With a Hamming window, we obtain the following frequency response: Refer to fig 10.9-1. 10.10 H(s) has two zeros at z1 = -0.1 and z2 = 8 and two poles p1,2 = -0.1 j3. The matched z-transform maps these into: -0.1T -0.01 z 1 = e = e =0.99 z 2 = e -8T =0 (-0.1+j3)Tj0.3 p 1 = e=0.99e-j0.3 p 2 =0.99e -1 1rz Hence, H(z)= ,w0 =0.3 r =0.99 12rcosw0z-1 + r2z-2 From the impulse invariance method we obtain

11 1 H(s)= + 2 s +0.1j3 s +0.1+j3 11 1 H(z)= + 1e-0.1Tej3Tz-1 1e-0.1Te-j3Tz-1 2 1rcosw0z-1 = 12rcosw0z-1 + r2z-2 314 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

---> mag(dB) ---> phase 3 2.5 2 1.5 1 0.5 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) 4 2 0 -2 -4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 ---> Freq(Hz) Figure 10.9-1: The poles are the same, but the zero is different. 10.11 ( u l)s Ha(s)= s2 ( u l)s + u l -1 21z s = z-1 T 1+

2 (1z-1)(1+z-1) T H(z) =( u l) 2 (2 )2(1z-1)2 +( u l)( )(1z-1)(1+z-1)+ u l(1+z-1)2 T T 2(a)(1z-2) = [4+2(a)+a]2(4a)z-1 +[4 2(a)+a]z-2 where a = uT, = lT In order to compare the result with example 10.4.2, let 3p wu = uT = 5 2p wl = lT = 5

0.245(1z-2) Then, H(z) = (example8.3.2) 1+0.509z-2 wu In our case, we have a = uT =2tan =2.753 2 wl = lT =2tan =1.453 2 315 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

By substituting into the equation above, we obtain 2.599(1z-2) H(z)= 10.599+5.401z-2 0.245(1z-2) = 1+0.509z-2 10.12 Let T =2 -1 1+z (a)H(z)= . y(n)= y(n 1)+x(n)+x(n 1) -1 1-z (b) 1 . = 0 2, Ha( )= H( )= , < 0 | | 2 (c)

w . 0 = w = p 2, |H(w)| = |cot | 6 H(w)=

2, -p<w< 0 2 (d) The digital integrator closelymatches the magnitude characteristics of the a nalog integrator. The two phase characteristics are identical. (e)Theintegrator hasapoleatw =0. To avoid overflow problems, we would have E[x(n )]=0, i.e., a signal with no dc component. 10.13 (a) -1)3 (1+z H(z)= A 1 z-2) (11z-1)(11z-1 + 224-1 (1+z-1)(1+2z+ z-2) 1 z-2) = A (11 z-1)(11 z-1 + 224H(z)|z=1 =1 3 111

. A = ,b1 =2,b2 =1,a1 =1,c1 = ,d1 = ,d2 = 64224 (b) Refer to fig 10.13-1 10.14 (a)There are only zeros, thus H(z)is FIR. (b) 4 Zeros: z1 = , 33 z2 = , 43

j 3 z3,4 = e 4 4

j 3 z5,6 = e 3 316

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) 3/64 3/64 x(n) + + z-1 z-1 + + + z-1 + + + z-1 z-1 + + y(n) -1 3 3 1 1/2 -1/8 + + z-1 -0.5 -1/2 2 11/4 1 y(n) Figure 10.13-1: z7 =1 1 Hence, z2 = z * 1 * z4 = z 3 1 z5 = z

* 3 * z6 = z 5 1 z1 = =1 z7 -1) and H(z)= z -6H(z Therefore, H(w)is linear phase. (c) Refer to fig 10.14-1 10.15 From the design specifications we obtain o =0.509 d = 99.995 41 fp = = 24 6 61 fs = = 24 4 317 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n) 25/12 y(n) -4/3 - 3/4 = -25/12 + + z-1z-1 z-1 + + z-1 z-1 z-1 + + z-1 + x(n) 25/12 y(n) -4/3 - 3/4 = -25/12 + + z-1z-1 z-1 + + z-1 z-1 z-1 + + z-1 + Assume t =1. Then, p and s . k Butterworth filter: Nmin Chebyshev filter: Nmin Elliptic filter: sina sin Nmin (4/3)2 + (3/4) 2 + 4 cos 2 60o = 481/144 Figure 10.14-1: wp =2tan 2 =2tanpfp =1.155 ws

=2tan 2 =2tanpfs =2 d = =196.5 o s = =1.732 p log. = =9.613 . N =10 logk cosh-1. = =5.212 . N =6 cosh-1k 1 = =0.577 . a =35.3o k 1 = =0.577 . =0.3o . k(sina) k(cos)

= . =3.78 . N =4 k(cosa) k(sin) 318 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.16 From the design specifications we have o = 0.349 d = 99.995 1.2 fp = 8 = 0.15 2 fs = 8 = 0.25 p = 2tan wp 2 = 1.019 s = 2tan ws 2 = 2 d . = = 286.5 o k = s = 1.963 p Butterworth filter: Nmin = log. logk = 8.393 . N = 9 cosh-1. Chebyshev filter: Nmin = cosh-1k = 4.90 . N = 5 Elliptic filter: Nmin = k(1 k) k( . 11 k2 ) . k( . 11 2 ) k(1 ) . N = 4

10.17 Passband ripple = 1dB . o =0.509 Stopband attenuation = 60dB . d =1000 wp =0.3p ws =0.35p wp p =2tan =1.019 2 ws s =2tan =1.226 2 d . = =1965.226 o s k = =1.203 p cosh-1. 8.277 = = =13.2 . N =14 Nmin cosh-1k 0.627 Special software package, such as MATLAB or PC-DSP may be used to obtain the fil ter coefficients. Hand computation of these coefficients for N =14 is very tedious.

319 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.18 Passband ripple = 0.5dB . o =0.349 Stopband attenuation = 50dB wp =0.24p ws =0.35p wp p =2tan =0.792 2 ws s =2tan =1.226 2 d . = =906.1 o s k = =1.547 p cosh-1. 7.502 = = =7.48 . N =8 Nmin cosh-1k 1.003 Use a computer software package to determine the filter coefficients. 10.19 (a)MATLAB is used to design the FIR filter using the Remez algorithm. We find th

at a filter of length M = 37 meets the specifications. We note that in MATLAB, the frequency scale is 1 normalized to of the sampling frequency. Refer to fig 10.19-1. 2 20 15 (b)d1 =0.02,d2 =0.01, .f = =0.05 100 100 0 0.2 0.4 0.6 0.8 1 1.2 |H(w)| 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 --> f Figure 10.19-1: With equation(10.2.94) we obtain v -20log10( d1d2)13

M = +1 34 14.6.f 320 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

With equation(10.2.95) we obtain D8(d1d2) =1.7371 f(d1d2) =11.166 D8(d1d2)f(d1d2)(.f)2 and M = +1 36 .f Note(10.2.95) is abetter approximation of M. (c) Refer to fig 10.19-2. Note that this filter does not satisfy the specifications. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 |H(w)| 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 --> f Figure 10.19-2: M=37 FIR filter designed by window method with Hamming window (d)The elliptic filter satisfies the specifications. Refer to fig 10.19-3. (e) FIR IIR order 37 5 storage 19 16 No. of mult. 19 16 10.20 (a) h(n) =0,1,2,3,4,5,4,3,2,1,0,... . 10 -n H(z)= h(n)z n=0 -8 -9 = z -1 +2z -2 +3z -3 +4z -4 +5z -5 +4z -6 +3z -7 +2z + z

H(w)= e -j9w[2cos4w +4cos3w +6cos2w +8cosw +5] (b)|H(w)| = |2cos4w +4cos3w +6cos2w +8cosw +5|. Refer to fig 10.20-1. 321 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 |H(w)| |H(w)| 0 0.05 0.1 0.15 0.2 0.25 --> f 0.3 0.35 0.4 0.45 0.5 Figure 10.19-3: 10.21 (a) dc gain: Ha(0) 3dB frequency: |Ha(j )|2 or a2 a2 + 2 c . c = = = = 1 1 2 1 2 a For all , only H(j8) ha(t) . e -t . t = = = = 0 1 e ha(0)= 1 e e -1 1 a

(b) h(n) H(z) H(w) = = = = ha(nT) e -nT u(n) 1 1e-Tz-1 1 1e-Te-jw H(0) 3dB frequency: |H(wc)|2 (1aT coswc)2 + (e -Tsinwc)2 = = = = H(w)|w=0 1 1e-T 1 2|H(0)|2 2(1e -T)2 322 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 5 10 15 20 25 --> magnitude 0 5 10 15 20 25 --> magnitude 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 --> Freq(Hz) Figure 10.20-1: Hence, wc =2sin-1(sinh aT ) 2 1 Since |H(w)|2 = e-2T 12e-Tcosw + 11 it oscillates between and e-T)2 (1e-T)2 (1+ but never reaches zero -T -1 h(t)= e = e 1 . t = aT 1 t is the smallest integer that is larger than T (c)

a H(z)= -1 21-z -1 + a T 1+z aT(1+z-1) = 2(1z-1)+aT(1+z-1) aT(1+z-1) = 2+aT +(aT 2)z-1 DC Gain: H(z)|z=1 =1 At z = -1(w = p),H(z) =0 1 since |Ha(j c)|2 = , we have c = a 2 -1 c wc =2tan T 2 =2tan-1aT2 2aT Let a = 2+aT 323 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1a -1 (1+a)z Then H(z) =1+ 1az-1 2 1a n-1 h(n) d(n)+(1+a)a u(n 1) = 2 1a h(0) = 2 h(n) h(0) . (1+a)a n-1 n = = = = 1 e 1 e ln a 1+a 1 lna ln(2-T 4 )1 ln(2-T 2+T ) 10.22 (a)

T T jwndw Hd(w)e hd(n) =

2p T T "Z-0.4 T = e 0.5 T 0.4 T jwndw jwndw + e 2p 2T T pn 2pn = np sin 2T sin 5T (b) Let hs(n) Then, h(n) = =

hd(n)w(n), 100 = n = 100(M = 101) hs(n 100) will be the impulse of the filter for 0 = n = 200 (c) Hd(w) = . ..... ..... 0, 0 =w =0.4 T e-j100w , 0.4 T =w =0.5 T 0, 0.5 T < w < 1.5 T -j100w 1.5 = w = 1.6 e, TT 1.6 <w = 2 0, TT 2 T w = k 200 pk = 100T Then, H(k) =0, 0 = k< 40 =0, 50 <k < 150

=0, 160 <k = 200 jk T H(k)= e, 40 = k = 50 -jkT = e, 150 = k = 160 H(w) will match Hd(w) at 201 points in frequency. The filter will contain large ripples in between the sampled frequencies. Transition values should be specified to reduce the ripples in both the passbands and the stopband. 324 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.23 (a) 5p wl = 12 wl l = tan (for T =2) 2 7p wu = 12 wu u = tan 2 Analog: lowpass to bandpass s2 + l u s . s( u l) Bilinear: Analog to digital -1 z 11z s . = z-1 z +1 1+

combine the two steps: 2 -1 1-z 1+z+ l u -1 s . -1 1-z ( u l) 1+z-1 -1)2 (1z-1)2 + u l(1+z = (1z-1)( u l) 1 Therefore, H(z)= i2 v -1)2 i h(1-z-1)2+ u l(1+z-1)2 h(1-z-1)2+ u l(1+z +2 +1 (1-z-2)( ul) (1-z-2)( ul)

(b) 7 u tan24 = =1.7 5 l tan24 u (1) =1.43 l u (2) =1.8 l u (3) =1.82 l u (4) =1.7 l filter(4) satisfiesthe constraint 10.24 (a) . 11 11 -1) (11 z-1)

(1+ z -12) 6 122122 H(z) =(1z ++ z-1 z-2 z-1 + z-2 . -13 -21 -3 -4 . 1+1z-1 + 1+ 1 2+z+ z+ z+ z -12) 22 = (1z z-5 61+z-1 + z-2 + z-3 + z-4 + 1 31 -1 -2 -3 -4) = (1z -6)(1z -1)(2+z + z + z + z 622 325

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

This filter is FIR with zeros at z =1,e j 6 ,e j 2 ,e j5 6 ,-0.5528 j0.6823 and 0.3028 j0.7462 (b) It is a highpass filter. (c) 1 31 -j3w -j4w) -j2w H(w)= (1e -j6w)(1e -jw)(2+e -jw + e + e + e 622 p 3p 3p H(0)= H( )= H( )= H( ) =0 666 2p 1 H( )=

62 4p H( ) =2 6 H(p) =2 10.25 (a) 900 fL = =0.36 2500 1100 fH = =0.44 2500 Refer to fig 10.25-1. (b) The ideal lowpass filter has a passband of-0.04 = f = 0.04. Hence, W(f)X(f) f f -0.4 0 0.360.40.44 -0.4 0.04 0.4 -0.04 0 V(f) f -0.04 0 0.04 Figure 10.25-1: . -j15w 1e-0.08p = w = 0.08p

Hd(w)= 0, otherwise Hence, . 0.08 1 -j15w jwndw hd(n)= ee 2p -0.08 326 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. 0.08 1 jw(n-15)dw = e 2p -0.08 2sin0.08p(n 15) = (n 15) h(n)= hd(n)wH(n) 2p(n 15) wH(n) =0.540.46cos 30 h(n)is the impulse response of the lowpass filter H(w) (c)Rx =1000 10.26 (a) M-1 y (n)= h(k)x(n k) k=0 8 E =[y(n)y (n)]2 n=0 8 M-1 =[y(n)h(k)x(n k)]2 n=0 k=0 By differentiating E with respect to each coefficient and setting the derivative s to zero, we obtain

M-1 h(k)rxx(kl)= ryx(l),l =0,1,...,M 1 k=0 8 where rxx(l)= x(n)x(n l) and n=0 8 ryx(l)= y(n)x(n l) n=0 (b) 8 M-1 E =[y(n)+w(n)h(k)x(n k)]2 n=0 k=0 By carrying out the minimization we obtain M-1 h(k)rxx(kl)= ryx(l)+rwx(l),l =0,1,...,M 1 k=0 8 where rwx(l)= w(n)x(n l) n=0 10.27 N-1 . N, n =0 j2kn/N x(n)= e= 0, 1 = n = N 1 k=0

327 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n)is a periodic sequence with period N. Hence, y(n)is also periodic with perio d N. Let p H2(z) =1+ akz k=1 and h2(n) -k = {1,a1,a2,...,ap} Then, h2(n)* y(n)= x(n),n =0,1,...,N 1 If p+1 = N, the N equations above are sufficient to determine a1,a2,...,ap and their orde r. If p+1 >N, it is not possible to determine the {ak} and the order p. 10.28 (1)The set of linear equations are: M-1 h(k)rxx(kl)= ryx(l),l =0,1,...,M 1 k=0 8. where rxx(l)= x(n)x(n l)and n=0 8 ryx(l)= y(n)x(n l) n=0 8 M-1 E = . n=-8 [y(n). k=0 (2)Refer to fig 10.28-1.

(3)Refer to fig 10.28-2. Total squared error h(k)x(n k)]2 error 8.28 8.26 8.24 8.22 8.2 8.18 8.16 8.14 8.12 8.1 8.08 8 9 1011121314 filter order Figure 10.28-1: (4)v(n)= y(n)+0.01w(n). Refer to fig 10.28-3. 328 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

fft of original h(n) M=11E=8.093 0.4 1 0.3 0.2 0.1 0 0 0 1000 2000 3000 0 1000 2000 3000 M=12E=8.084 M=13E=8.101 0.5 0.5 0.4 0.4 -->|H(w)| -->|H(w)| -->|H(w)| -->|H(w)| 0.5 0.1 0.1 0 0 0 1000 2000 3000 0 1000 2000 3000 Figure 10.28-2: 10.29 (a)Since d(n k)=0 exceptfor n = k, equation(1) reducesto h(n)= -a1h(n 1)a2h(n -

2)... aNh(n N)+bn,0 = n = M (b) Sinced(n k)=0 exceptfor n = k, equation(1) reducesto h(n)= -a1h(n 1)a2h(n 2)... aNh(n N),0n>M (c)We usethelinearequationgivenin(b)tosolveforthe filterparameters{ak}. Then we use values for the {ak} inthelinearequation fivenin(a) andsolvefortheparameters {bk}. 10.30 2 Hd(z)= 11z-1 2 We can see that by setting M =0 and N =1 in PM bkz-1 k=0 H(z)= 1+PN k=1 akz-1 we can provide a perfect match to Hd(z)as given in b0 H(z)= . 1+a1z-1 329 0.3 0.3 0.2 0.2

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

fft of original h(n) M=8 E=8.42 M=9 E=8.305 M=8 E=8.42 M=9 E=8.305 0 0.5 1 1.5 -->|H(w)| 0 0.5 0 0.2 0.4 0 1000 2000 3000 0 1000 2000 3000 0 1000 2000 3000 M=10 E=8.228 M=11 E=8.228 M=12 E=8.221 0 0.2 0.4 0 0.2 0.4 0 0.5 0 1000 2000 3000 0 1000 2000 3000 0 1000 2000 3000 M=13 E=8.238 M=14 E=8.24 0 0.5 0 0.5 0 1000 2000 3000 0 1000 2000 3000 Figure 10.28-3: With d(n)as the input to H(z), we obtain the output h(n)= -a1h(n 1)+b0d(n). For n>M =1, we have h(n)= -a1h(n 1) or, equivalently, hd(n)= -a1hd(n 1). Substituting for hd(n), we obtaina1 = 1 . Tosolvefor b0,we usetheequationgivenin10.29(a) 2 with h(n)= hd(n), 1 hd(n)= hd(n 1)+b0d(n).

2 For n =0 this equation yields b0 =2. Thus 2 H(z)= . 11 z-1 2 (b) Hd(z)= H(z) 330 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.31 (a) N M X. h(n)= k=1 akh(n k)+ k=0 bkd(n k) n = 0 (b) Based on (a) N

hd(n)= akhd(n k) n>M k=1 8 X i2 hd(n)h d(n) e1 = M+1 8 . #2 N = hd(n)akhd(n k) M+1 k=1 By differentiating with respect to the parameters {ak}, we obtain the set of lin ear equations of the form N akrhh(k,l)= -rhh(l,0) l = 1,2,. . . ,N k=1

where, 8 . rhh(k,l)= hd(n k)hd(n l) n=1 8 . = hd(n)hd(n + kl)= rhh(kl) n=0 The solution of these linear equations yield to the filter parameters {ak}. (c) We can find the least-squares solution for{bk} from the minimization of 8 . #2 M

hd(n)e2 bkv(n k) = n=0 k=0 Thus we obtain a set of linear equations for the parameters {bk}, in the form M bkrvv(k,l)= rhv(l) l = 0,1,. . . ,M k=0 where 8 . rvv(k,l)= v(n k)v(n l)

n=0 8 . rhv(k)= h(n)v(n k) n=0 331 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

10.32 (a) y(n) =1.5198y(n 1)0.9778y(n 2)+0.2090y(n 3) +0.0812x(n)+0.0536x(n 1)+0.0536x(n 2)+0.0812x(n 3) hd(n)can be found by substituting x(n)= d(n). Fig 10.32-1 shows the hd(n). 0 5 101520253035404550 n Figure 10.32-1: -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25 0.3 h d (n) (b)Thepolesand zerosobtained using Shanks method arelistedinTable10.32. Themagnitu de responsefor each casetogetherwith the desiredresponseisshowninFig.10.32-2. Thefr equency responsecharacteristicsillustratethat Shanks methodyieldsverygooddesignswhenthenum ber of poles and zeros equals or exceeds the number of poles and zeros in the actual filter. Thus the inclusion of zeros in the approximation has a significant effect in the resultin g design. Filter Order Poles Zeros N=3 0.5348 M=2 0.6646j0.4306 -0.2437j0.5918 N=3 0.3881 -1 M=3 0.5659j0.4671 0.1738j0.9848 N=4 -0.00014 -1 0.388 M=3 0.566j0.4671 0.1738j0.9848 332 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-90 -80 -70 -60 -50 -40 -30 -20 -10 0 Magnitude (dB) Desired response N=3, M=3 N=4, M=3 N=3, M=2 -90 -80 -70 -60 -50 -40 -30 -20 -10 0 Magnitude (dB) Desired response N=3, M=3 N=4, M=3 N=3, M=2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Normalized Frequency (xp rad/sample Figure 10.32-2: 333 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

334 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 11 11.1 (a)Let the corresponding baseband spectrum be called Xb( ). Then 1 Xa( )= [Xb( 2000p)+Xb( +2000p)] 2With frequencies normalized to Fx,

' w = Fx . The sequence x(n)has DTFT 8 ' X(w ' )= Xa(w 2pq) q=-8 8 '' =[Xa(w 0.8p 2pq)+Xb(w +0.8p 2pq)] q=-8 1 modulation by cos(0.8p) causes shifts up and down by 0.8p (and scaling by ) of each 2 |X(w )|

Assumes peak of X(.) normalized to unity b 0.5 X (w ) shifted to 0.8p period 2p b w -p -0.8p 0 0.8p p b=0.16p Figure 11.1-1: componentinthespectrum. Referto fig11.1-1. IdealLPF preservesonly the basebandsp ectrum (ofeach period). Refer to fig 11.1 335 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0.5 1 |W(w )| 0.5 H(w ) period 2 p -p -0.4p 0.1p 0.4p p w b Figure 11.1-2: ''

D Fx = ' 10w . Note that Thedownsamplingproducesthe figurein fig11.1,where w = = Fy ' there is no aliasing in the spectrum |Y(w '' )| because the decimated sample rate, in terms of w , is 210 > 0.04p. (b) The assumed spectral amplitude normalization in fig 11.1-1 implies that the analog FT (magnitude spectrum)of xa(t)is(referto fig11.1-4). The given sample rate is identical to Fy above, Fy = 250Hz. The DTFT of samples taken 1 Ty q at this rate is Y ( ) Xa( q y) where y =2pFy. On a scaled frequency axis

= = 1 Ty q

Y(w

'' '' ) '' Xa(w q2p). Consequently Ty = w = , Fy |V(w )| w period 2 p p 0.08p 0.8p p 0.1 period 2 w |Y(w )| = 0.1 |V(0.1w p Figure 11.1-3: 336 y(n)= y(n).

)|

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

|Xa ( W) | Tx /2 Wc W -Wc 400p Figure 11.1-4: 11.2 1 (a)X(w)= -jw) (1-ae X(w ' 1 1 (b) After decimationY(w ' )= )= jw ' 22 2(1-ae 2 ) (c) -jw2n DTFT {x(2n)} = x(2n)e n -jw ' n = x(2n)e n = Y(w ' ) 11.3 (a)Refer to fig 11.3-1

(b) z -1 + x(n) F x 1/2 y(m) Fy = Fx Figure 11.3-1: 337 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

'

w ' '' Let w = ,w = = Fx Fy 2 . n -jw '' . 1 n 1 n +1 -jw '' '' ) n n Y(w = x( )e + [x( )+x( )]e 2222 n even n odd . -jw '' 1 . -jw '' 2p (2q+1) = x(p)e + [x(q)+x(q+1)]e 2 pq 1 -jw '' '' j2w '' )+

= X(2w e [X(2w '' )+eX(2w '' )] 2 '' ] = X(2w '' )[1+cosw 1, 0 =|w ' |= 0.2p ' ) X(w = 0, otherwise 1, 0 =|2w '' |= 0.2p '' ) X(2w = 0, otherwise 1, 0 =|w '' |= 0.1p = 0, otherwise . '' 1+cosw , 0 =|w '' |= 0.1p '' ) Y(w = 0, otherwise (c) Refer to fig 11.3-2 X(.) 0 0.7 0.9 1.1 1.3 2 w p p p pp p 0 0.35 p 0.45 p 0.55 p 0.65 pp w Figure 11.3-2: . '' 1+cosw , 0.35p =|w '' |= 0.45p

. '' ) Y(w = or0.55p =|w '' |= 0.65p . 0, otherwise 11.4 ' '' D (a)Let w = ,w = . Refer to fig 11.4-1 Fx Fx Let x '' (n)be the downsampled sequence. 338 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

p |X(w )| 1 -w m w m w p |X(w )| 1 -w m w m w p |X (w )| 1/D -Dw m Dw m w Figure 11.4-1: '' (n) x = x(nD) '' '' (w '' ) 1 w X = X( ) DD '' (w '' )[x '' (n) As long as Dw ' = p, X(w ' ) [hence x(n)] can be recovered from X = x(Dn)] m using interpolation by a factor D: X(w ' )= DX '' (Dw ' ) ' 2 ' = 2 ' The given sampling frequency is w = . The condition Dw ' = p . 2w = w sD mmDs

(b)Letxa(t)be the ralanalog signalfrom which samples x(n)were taken at rate Fx. There exists '' t a signal, say x (t ' ), suchthat x (t ' )= Xa( ). x(n)maybe consideredtobethesamplesof x ' (t ' ) aaTx '' (n)' (t ' )'' fx 1 taken at rate fx = 1. Likewise x = x(nD)are samples of x taken at rate f = = xDD. From sampling theory, we know that x ' (t ' )can be reconstructed from its sampl es x '' (n)as long as it is bandlimited to fm = 1 , or wm = , which is the case here. The reconstruction formula 2DD is '' (k)hr' x ' (t ' )= x (t kD) k where

t ' ) sin( D hr(t ' )= ( t ') D Refer to fig 11.4-2 ' Actually the bandwidth of the reconstruction filter may be made as small as w , or as large m 2 '

as w , so hr may be Dm ' sin(wt ' ) c hr(t ' )= (w ' t ') c '' ' where w = w = 2 w . In particular x(n)= x ' (t ' = n)so mcD m x(n)= x(kD)hr(n kD) k 339 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

|X(w (w ) Hr p/D w w mm 0

)| period2 p/D

2 p/D - w 2 p/D

Figure 11.4-2: (c) Clearlyif we define . x(p), if p is an integer multiple of D v(p)= 0, other p then, we may write 11.4 as x(n)= v(p)hr(n p) p so x(n)isreconstructedas(see fig11.4-3) x (n)=x(kn) D v(n) x(n) h r(n) Figure 11.4-3: 340 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.5 (a) '' 2 Let w = , w = Fx Fx Xs(w) = . xs(n)e -jwn n = . x(2m)e -jw2m m = 1 . X(w 2p q) 2 2 q 1 . = X(w pq) 2 q To recover x(n)from xs(n): see fig 11.5-1 (b) X (w) s period p -p -2p/3 -p/3 2p/3 w p/3 p 1/2 x (n) 2 v(n) h(n) rs x(n)

where H r (w) 1 -p/2 p/2 w Figure 11.5-1: ' Recall w = 2w Xd(w ' ) = . xd(n)e -jw ' n n = . xs(n)e -jw ' n 2 n even = . xs(n)e -jw ' n 2 n 341 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

since xs(n)=0 when n odd ' w = Xs( ) 2 see fig 11.5-2 No information is lost since the decimated sample rate still exceeds twice the b andlimit of 1/2 period 2 p -p 2p/3 p w Figure 11.5-2: the original signal. 11.6

A filter of length 30 meets the specification. The cutoff frequency is wc = and the coefficients 5 are given below: h(1) h(2) h(3) h(4) h(5) h(6) h(7) h(8) h(9) h(10) h(11) h(12) h(13) h(14) h(15) 342 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o = = = = = = = = = = = = = = = h(30)=0.006399 h(29)= -0.01476 h(28)= -0.001089 h(27)= -0.002871 h(26)=0.01049 h(25)=0.02148 h(24)=0.01948 h(23)= -0.0003107 h(22)= -0.03005 h(21)= -0.04988 h(20)= -0.03737 h(19)=0.01848 h(18)=0.1075 h(17)=0.1995 h(16)=0.2579

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

pk(n)= h(n + k),k =0,1,... correspondingpolyphase filterstructure(seefig11.6-1) + p0 (n) + p (n) 1 y(n) x(n) p (n) 4F x F y = Fx /D Figure 11.6-1: 11.7

A filter of length 30 meets the specification. The cutoff frequency is wc = and the coefficients 2 are given below: h(1) h(2) h(3) h(4) h(5) h(6) h(7) h(8) h(9) = = = = = = = = = h(30)=0.006026 h(29)= -0.01282 h(28)= -0.002858 h(27)=0.01366 h(26)= -0.004669 h(25)= -0.01970 h(24)=0.01598 h(23)=0.02138 h(22)= -0.03498

h(10) = h(21)= -0.01562 h(11) = h(20)=0.06401 h(12) = h(19)= -0.007345 h(13) = h(18)= -0.1187 h(14) = h(17)=0.09805 h(15) = h(16)=0.4923 pk(n)= h(2n + k),k =0,1; n =0,1,..., 14 343 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

correspondingpolyphase filterstructure(seefig11.7-1) p0 (n) /DF y = FxF x p (n) 1 y(n) x(n) Figure 11.7-1: 11.8 TheFIRfilterthatmeetsthespecificationsof thisproblemisexactly thesameasthatinPro blem 11.6. Its bandwidth is 5. Its coefficients are g(n,m)= h(nI +(mD)I) mD = h(nI + mD []I) I 5m = h(2n + 5m 2[ 2 ]) g(0,m) g(1,m) = = {h(0),h(1)} {h(2),h(3)} . . . g(14,m) = {h(28),h(29)} A polyphase filter would employ two subfilters, each of length 15 p0(n)= {h(0),h(2),...,h(28)} p1(n)= {h(1),h(3),...,h(29)} 11.9 (a) x(n)= {x0,x1,x2,...} D = I =2. Decimation first 344

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

y2(n)= x0,0,..., 0,xdk,0,..., 0,x2dk,... .. ik-1 ik-1 .. . .. . .. z1(n)= x0,0,..., 0,x1,0,..., 0,x2,0,..., 0,... .. ik-1 ik-1 ik-1 .. . .. . .. y1(n)= x0,0,..., 0,xd,0,..., 0,... .. d-1 d-1 .. .. .. = h0,0,..., 0,H1,0,..., 0,h(2),... . . D-1 D-1 nD-1

z2(n)= {x0,x2,x4,...} y2(n)= {x0,0,x2,0,x4,0,...} Interpolation first z1(n)= {x0,0,x1,0,x2,0,...} y1(n)= {x0,x1,x2,...} so y2(n) 6y1(n) = (b) supposeD = dk and I = ik and d,i are relatively prime. x(n)= {x0,x1,x2,...} Decimation first z2(n)= {x0,xdk,x2dk,...} . .. . ..

h(kD)x(n kD)

h(k)x(n k) Interpolation first Thus y2(n)= y1(n)iff d = dk or k =1 which means that D and I are relatively prim e. 11.10 (a)Refer to fig 11.10-1 y1(n)= h(n)* w1(n) = h(n)* x(nD) 8 = h(k)x[(n k)D] k=0

0 D 2D H(z D)= ...h(0)z + h(1)z + h(2)z + ... H(z D) . h (n) so w2(n) = k=0 n = k=0 n = h(k)x(n kD) k=0 345 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

D w1(n) H(z) x(n) y(n) 1 H(zD) w2(n) D y2(n) x(n) Figure 11.10-1: y2(n)= w2(nD) n = h(k)x(nD kD) k=0 n So y1(n) = = k=0 h(k)x[(n k)D] y2(n) (b) 8 . w1(n)= h(k)x(n k) k=0 y1(n)= w1(p),n = pI(p an integer) =0, other n w2(n)= x(p),n = pI =0, other n Let h (n)be the IR corresponding to H(zI) 8 X

h(k)w2(n k)

y2(n) = k=0 8 . = k=0

h(kI)w2(n kI) 8 . = h(k)w2(n kI) k=0 for n = pI 8 . y2(n)= h(k)w2((pk)I) k=0 346 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

8 = h(k)x(pk) k=0 = w1(p)(see above) for n 6= pI 8 y2(n)= h(k).0 =0 k=0 so we conclude y1(n)= y2(n) 11.11 (a) -2n -2n-1 H(z)= h(2n)z + h(2n +1)z nn = h(2n)(z 2)-n + z -1 . h(2n +1)(z 2)-n nn = H0(z 2)+z -1H1(z 2) -n Therefore H0(z)= h(2n)z n -n H1(z)= h(2n +1)z n (b) -nD -nD-1 H(z)= h(nD)z + h(nD +1)z + ... nn -nD-D+1 + h(nD + D1)z n D-1

. -k . D)-n = zh(nD + k)(z k=0 n -n Therefore Hk(z)= h(nD + k)z n (c) 1 H(z)= 1az-1 8 n -n = az n=0 8 2n -n H0(z)= az n=0 1 = 1a2z-1 8 2n+1 -n H1(z)= az n=0 a = 1a2z-1 347

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.12 The output of the upsampler is X(z2). Thus, we have Y1(z)= 1 X(z)H1(z1/2)+X(z)H1(z1/2W1/2). 2 1 . 1/2W1/2). = H1(z1/2)+H1(zX(z) 2 = H2(z)X(z) 11.13 (a)Refer to Fig. 11.13-1 for I/D =5/3. DTFT[x(n)] 5F F x x Filter

5F F x x 3FyFy2Fy(1/2)min(Fx,Fy) DTFT[y(m)] Figure 11.13-1: (b) Refer to Fig. 11.13-2 forI/D =3/5. 11.14 (a)The desired implementation is given in Fig. 11.14-1

(b) The polyphase decomposition is given by -1)5 Hk(z) = (1+z -4 -5 = 1+5z -1 +10z -2 +10z -3 +5z + z -2 -1 = 1+10z -2 +5z -4 +(5+10z + z -4)z -1 = P0(z)+P1(z)z 348 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

DTFT[x(n)] 3F Fx x Filter (1/2)min(Fx,Fy)

3F x Fy5FyDTFT[y(m)] Figure 11.13-2: 11.15 (a) N-1 H(z)= z -nPn(z N) n=0 where 8 -k Pn(z)= h(kN + n)z k=-8 Let m = N 1n. Then N-1 H(z) = . z -(N-1-m)PN-1-m(z N) n=0 N-1 = . z -(N-1-m)Qm(z N) n=0 (b) ( )1 15

+ z 2 ( )1 15 + z 2 ( )1 15 + z 2 Figure 11.14-1: 349 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

y(n) P (z )N0 z-1 z-1 + + + P (z )N1 N-2 NP (z ) x(n) y(n) x(n) N N P (z ) P (z ) N-1N-1 N 0 + + + z-1 z-1 P (z )NP (z ) 1 N-2 NP (z ) Figure 11.15-1: Type 1 Polyphase Decomposition x(n) y(n) N-1 + + z-1 z-1 + 0Q (z )N NQ (z Q (z )1 N N ) Q (z )N-2 Figure 11.15-2: Type 2 Polyphase Decomposition 350 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.16 Q (z ) Q (z )N + z-1 z-1 + 3 3 3 N x(n) Q (z )1 y(n) 2 N 0 Figure 11.16-1: 11.17 D1 = 25,D2 =4 F0 F0 = 10 kHz ,F1 = =400 Hz D1 Passband 0 = F = 50 Transition band 50 <F = 345 Stopband 345 <F = 5000 F1 F2 = =100 Hz D2 Passband 0 = F = 50 Transition band 50 <F = 55 Stopband 55 <F = 200 For filter 1, d1 =

0.1 =0.05,d2 =10-3 2 34550 .f = =2.95x10-2 10,000 -10logd1d2 13

M1 = +1 =71 14.6.f For filter 2, d1 =0.05,d2 =10-3 5550 .f = =7.5x10-3 400 -10logd1d2 13

M2 = +1 275 14.6.f The coefficients of the two filters can be obtained using a number of DSP softwa re packages. 351 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.18 To avoid aliasing Fsc = Fx . Thus D = I = 50. 2D Single stage d1 =0.1,d2 =10-3 6560 .f = =5x10-4 10,000 -10logd1d2 13

M1 = +1 3700 14.6.f Two stages D1 = 25,D2 =2 I1 =2,I2 =25 10,000 stage 1:F1 = =400 25 Passband 0 = F = 60 Transition band 60 <F = 335 Stopband 335 <F = 5000 10-3 d1 =0.1,d2 = 4 .f =2.75x10-2 M 1 =84 400 stage 2:F2 = =200 2 Passband 0 =

F = 60 Transition band 60 <F = 65 Stopband 65 <F = 100 10-3 d1 =0.1,d2 = 4

.f =0.1875 M2 =13 Use DSP software to obtain filter coefficients. 11.19 b+(n)is nonzerofor0= n = 2N2 with N even. Let c(n)= b+[n(N1)]. So c(n)is nonzero for -(N 1) = n = N 1. From(11.11.35) ae-jw(N-1) B+(w)+(-1)N-1B+(w p)= or B+(z)+(-1)N-1B+(-z)= az-(N-1) Therefore, C(z)z -(N-1) +(-1)N-1C(-z)(-z)-(N-1) = az-(N-1) or C(z)+C(-z)= a c(n)+c(-n)= ad(n) when n 60c(n)= -c(-n) = when n is odd c(n)= -c(-n) when n is even but n 6=0,c(n) =0 (half-band filter) a when n =0,c(n)= 2 352 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.20 one stage: d1 =0.01,d2 =10-3 10090 .f = =10-3 10,000 -10logd1d2 13

M = +1 2536 14.6.f two stages: F0 =2 105 Hz I1 =1,I2 =2 F0 F1 = =2 104 Hz I1 Passband 0 = F = 90 Transition band 90 <F = 19,900 19,90090 Therefore .f = =0.09905 2 105 d1 and d11 = ,d12 = d2 2 -10logd1d2 13

M1 = +1 29 14.6.f F1

F2 = =1 104 Hz I2 Passband 0 = F = 90 Transition band 90 <F = 9,900 9,90090 Therefore .f = =0.4905 2 104 d1 and d21 = ,d22 = d2 2 -10logd1d2 13

M2 = +1 7 14.6.f 11.21 Suppose the output of the analysis section is xa0(m)and xa1(m). After interpolat ion by 2, they become y0(m)and y1(m). Thus . xak(m),m even k =0,1 2 yk(m)= 0,m odd The final output is z(m)= y0(m)* 2h(m)+y1(m)* [-2(-1)mh(m)] when m is even, say m =2j, z(m)= z(2j) =2y0(m)* h(m)2y1(m)* h(m)

=2 y0(k)h(m k)2 y1(k)h(m k) kk =2 y0(2l)h(2j2l)2 y1(2l)h(2j2l) ll 353 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

=2 xa0(l)h[2(jl)]2 xa1(l)h[2(jl)] ll = 2[xa0(j)xa1(j)]* h(2j) = 2[xa0(j)xa1(j)]* p0(j) In the same manner, it can be shown that z(2j+1) =2[xa0(j)+xa1(j)]* p1(j) 11.22

Refer to fig 11.22-1, where hi(n)is a lowpass filter withcutoff freq. Ii . Aftertransposition(refer I1 h1(n) I h(n) L L Interpolator 1 Interpolator L Figure 11.22-1: I = I1I2 ...ILL-stage interpolator to fig 11.22-2). As D = I, let Di = IL+1-i, then D = D1D2 ...DL. Refer to fig 11 .22-3 Obviously, this is equivalent to the transposed form above. 354 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

h(n) L h 1(n)I L I 1 h(n) L h 1(n)I L I 1 Figure 11.22-2: h (n) L h1(n)DL D1 Figure 11.22-3: L-stage decimator 355 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.23 k 1 I 1 I Suppose that output is y(n). Then Ty = Tx. Fy = = = Fx. Assume that the lowpass ITy kTx k filter is h(n)of length M = kI (see fig 11.23-4) 1 2 K buffer length K input buffer length K x(n) F x output buffer length I y(n) F y = ( I/k) Fx K-1 n=0 + g(n,0) g(n,1) g(n,I-1) n=0,1, ..., K-1 n=0,1, ..., K-1 n=0,1, ..., K-1 coefficient storage Figure 11.23-4: 11.24 (a) for any n = lI + j (0= j= I 1) I-1 I-1

pk(n k) = pk(lI + jk) k=0 k=0 = pj(lI) = pj(l) = h(j+ lI) = h(n) I-1 Therefore, h(n)= pk(n k) k=0 356 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) z-transform bothsides I-1 -k H(z)= zpk(z) k=0 (c) I-1 1 . 2l(n-k) jI h(n)ez n-k I I nl=0 I-1 1 X. j2lm -m = h(k+ mI)ez I ml=0 -m = h(k+ mI)z m -m = pk(m)z m = pk(z) 11.25

(a)Refer to fig 11.25-1. (b) --> magnitude spectrum of x(n) spectrum of y(n) 1 0.8 0.8 0.7 --> magnitude 0.6

0.6 0.4 0.5 0.2 0 0.4 02468 02468 --> w --> w Figure 11.25-1: p Bandwidth = 3 p cut off freq = 2 sampling freq of x(n) =2p 2p sampling freq for the desired band of frequencies =

= p 2 2p Therefore, D = =2 2 (c) Refer to fig 11.25-2. (d) Refer to fig 11.25-3. 357 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

-->x(n) x(n) |X(w)| 40 1200 1000 --> magnitude 30 800 600 400 20 10 200 0 0 0 500 1000 1500 0 500 1000 1500 ---> n Figure 11.25-2: spectrum of s(n) 0 100 200 300 400 500 600 700 800 900 1000 --> magnitude 0 200 400 600 800 1000 1200 Figure 11.25-3: 358 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.26 N-1 Q (z )N + + z-1 z-1 + y(n) I I I I N N N x(n) Q (z ) Q (z ) Q (z )1 N-2 0 Figure 11.26-1: 11.27 N-1 -nPn(z N) H0(z) = z n=0 where 8 Pn(z)= h0(kN + n)z -k , 0 = k = N 1 k=0 Then, -j2k/N)k

Hk(z)= H0(ze = H0(zw N) -j2/N where wN = e. (a) N-1 -l -kl NkN N ) Hk(z) Pl(z = zw w N l=0 N-1 = z -l -kl wPl(z N),k =0,1,...,N 1 N l=0 Therefore, Hk(z),0 = k = N 1 can be expressed in matrix form as . P0(zN) z-1P1(zN) . . . z-1P1(zN) ... . -k -2k -(N-1)k

Hk(z)=1 w w ... w N N N 359 . ... .

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b)From part(a),we have . .. . . 11 1 1 P0(z N) H0(z) -1 -2 -(N-1) -1P1(zN) H1(z) . . . 1 w ... ... = ... w w ... ... z N N N

... . ... . ... . . . . -(N-1) -2(N-1) -(N-1)(N-1) N) HN-1(z)1 ww wz-1P1(z NN N . . . ... = NW-1 ... N) z-1P1(zN) P0(z . . . z-1P1(zN) ... where W id the DFT matrix. (c) IDFT N 0 z-1 z-1 P (z )N1 N-2

N-1 N NP (z ) P (z ) x(n) y (n) y (n) 1 N-2 N-1y (n) y (n) 0 N-point P (z ) Figure 11.27-1: (d) y (n) 0 DFT y (n)1 N-2y (n) N-1y (n) P (z )N0 P (z )N1 + + + z-1 z-1 N-point v(n) P (zN ) N-2 P (zN) N-1 Figure 11.27-2: 360 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

11.28 -3 H0(z)=1+z -1 +3z -2 +4z (a) k Hk(z)= H0(zw 4), 1 = k = 3 -j2k/4) = H0(ze Then, -2 -3 H1(z) =1+jz-1 3z + j4z -3 H2(z) =1jz-1 +3z -2 4z -2 -3 H3(z) =1jz-1 3z + j4z Note that the impulse response hk(n)are complex-valued, in general. Consequently , |Hk(w)| is not symmetric with respect to w =0. (b) Let us use the polyphase implementation of the uniform filter bank. We have 8 Pl(z)= h0(l+3n)z -n , l =0,1,2,3 n=0 This yields P0(z) =1, P1(z) =1, P2(z) = 3, and P3(z) = 4. By using the results i n Problem 11.27, we have the equation for the synthesis filter bank as .

.. . . . P0(z 4) H0(z) 11 1 1 -1P1(z4) H1(z) H2(z) 1 j -1 -j 1 -11 -1 .. . .. . = .. . .. . .. . .. . z -2P2(z4) z H3(z)1 -j -1 jz-3P3(z4)

. . . . . . 11 1 1 1 1 -1 -1 1 j -1 -j 1 -11 -1 .. . .. . .. . .. . = 4W-1 .. . .. . z z =

-2 -2 3z 3z -3 -3 1 -j -1 j 4z4zwhere W denotes the DFT matrix. Thus, we have the analysis filter bank given in fig 11.28-1. (c) The synthesis filter bank in fig. 11.28-2 11.29 -6 H(z)= -3+19z -2 +32z -3 +19z -4 3z (a) -1) 4 3z 6 H(z = -3+19z 2 +32z 3 +19z z -6H(z -1)= -3z -6 +19z -4 +32z -3 +19z -2 3 = H(z) Therefore, H(z-1) and H(z) hve roots that are symmetric, such that if zi is not a root, then 1/zi is also a root. This implies that H(z)has linera phase. 361 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

y (n) -1 + z-1 z-1 1 y (n) y (n) 1 0 IDFT x(n) 2 3 4-point 4 0P (z )4 P (z )4P (z )4P (z )4 2 3y (n) z 1 y (n) y (n) 1 0 IDFT x(n) 2 3 4-point 4 0P (z )4 P (z )4P (z )4P (z )4 2 3y (n) z Figure 11.28-1: y (n) 0 v(n) y (n) 3 + z-1 + + z-1 z-1 DFT 1 y (n)1 0P (z )NP (z )N2

3 N NP (z ) P (z ) 4-point 2y (n) Figure 11.28-2: (b) We mayexpress H(z)as: -3 -3 H(z)= z -3z 3 +19z 1 +32+19z -1 3z Thus,wehavethe coefficients: 32,n =0 h(2n)= 0,n =6 Therefore, H(z)is a half-band filter. 362 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(c) 70 60 50 angle(H(w)) |H(w)| 40 30 20 10 0 -4-3-2-1 0 1 2 3 4 w 10 5 0 -5 -10 -4-3-2-1 0 1 2 3 4 w Figure 11.29-1: 11.30 -1 H0(z)=1+z (a) 8 Pl(z)= h0(l+2n)z n=0 8 -n P0(z)= h0(2n)z -n = 1

n=0 8 P1(z)= h0(l+2n)z -n = 1 n=0 363 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) H1(z) = P0(z 2)z -1P1(z 2) = 1z -1 x(n) 1 1 z-1 + + 2 2 Figure 11.30-1: Anaylsis section (c) G0(z) G1(z) = = P0(z 2)+z -1P1(z 2) = P0(z 2)z -1P1(z). 1+z -1 = 1+z -1 -1 2 2 1 1 + + + + 1 1 z-1 + 2 2 x(n)^ x(n) z Figure 11.30-2: QMF in a polyphase realization (d) For perfect reconstruction, Q(z)= 1[H0(z)G0(z)+H1(z)G1(z)]= Cz-k 2 where C is a constant. We have -1)2. -1

Q(z)= 1 (1+z -1)2 (1z =2z 2 11.31 (a) . .. -1 -2 . H0(z) 1+z+ z -1 -2 H(z)= . H1(z) . = . 1z+ z. = P(z 3)a(z) -2 H2(z)1z .. 1 -1 where a(z)= . z.. Then -2 z . -1 -2 .. ... 1+z+ zP00(z3) P01(z3) P02(z3)1 -1 -2 -1 . 1z+ z. = . P10(z3) P11(z3) P12(z3) ..

z. -2 -2 1zP20(z3) P21(z3) P22(z3) z364 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

.. 11 1 Clearly, P(z3)= . 1 -11 . 10 -1 (b) The synthesis filters are given as -3Qt -1) G(z)= z (z 3)a(z -1 where Q(z)= Cz-k [P(z)].But .. 11 2 [P(z)]= -1 1 . 2 -20 . 4 11 -2 By selecting C =4 and k =1, we have .. 11 2 Q(z)= z . 2 -20 . 11 -2 Therefore, ... ... G0(z) 112 1 -2 -1 . G1(z) . = z . 2 -20 .. z. -2

G2(z) 11 -2 z . -1 -2 . 1+2z+ z -1 -2 = . 12z+ z. -1 -2+2z (c) x(n) v(n) 3P(z) Q(z) 3 z-1 3 3 z-1 + z-1 z-1 + 3 3 Figure 11.31-1: 365 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

366 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 12 12.1 (a) 25 ..xx(z)= 11 z-2) (1z-1 + z-2)(1z-1 + 22 1 H(z)= 1z-1 + 1 z-2 2and s2 = 25 w 1 so x(n)= x(n 1)x(n 2)+w(n) 2 -11 -2 (b) The whitening filter isH-1(z)=1z+ z 2 12.2

27 (1-1 z1)(1-1 z) 33 (a)..xx(z)= 2 (1-1 z1)(1-1 z) 22 For a stable filter, denominator(11 z1) must be chose. However, either numerator factor 2(11 z1) (1-1z) 3 may be used. H(z)= or 3 (1-1z) 2 (11z1) 2 [min.pk.] (b) Must invert the min. pk. filter to obtain a stable whitening filter. (11 z1) 2 H-1(z)= (11 z1) 3 12.3 (a) -1 1+0.9z H(z)= 11.6z-1 +0.63z-2 11.6z-1 +0.63z-2 whitening filter, H-1(z)= 1+0.9z-1 zeros: z =0.7 and 0.9

pole: z = -0.9 367 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) ..xx(w)= s2 H(w)H(-w) w |1+0.9e-jw|2 s2 = w |11.6e-jw +0.63e-2jw|2 12.4 13 5 1 -1 -2 -3 A(z) =1+ z + z + z 24 8 3 1 k3 = 3 15 13 -1 -2 -3 B3(z)= + z + z + z 38 24 1 k3 =

2 13 -1 -2 B2(z)= + z + z 28 A2(z)k2B2(z) A1(z)= 1k22 1 -1 = 1+ z 4 1 k1 = 4 12.5 1 -1 -2 A2(z) = 1+2z + z 3 1 -1 -2 B2(z)= +2z + z 3 1 k2 = 3 A2(z)-

k2B2(z) A1(z)= 1k22 3 -1 = 1+ z 2 3 k1 = 2 12.6 (a) 1 -1 A1(z) =1+ z 2 1 -1 B1(z)= + z 2 -1 A2(z)= A1(z)+k2B1(z)z 368 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

= 1+ 1 z -1 1 z -2 3 3 B2(z) = 1 3 + 1 3 z -1 + z -2 H(z)= A3(z) = A2(z)+k3B2(z)z -1 = 1+z -3 The zeros are at z = -1,e j 3 Refer to fig 12.6-1 1 Figure 12.6-1: (b) If k3 = -1, we have H(z)= A3(z) = A2(z)B2(z)z -1 = 1+ 2 z -1 2 z -2 z -3 3 3v 5 11 The zeros are at z = -1,6 j 6 (c) If|kp| =1, the zeros of H(z)= Ap(z)are on the unit circle. Refer to fig 12.6-2. 369 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

unit circle unit circle Figure 12.6-2: 12.7 -1 A1(z) = 1+0.6z -1 B1(z) =0.6+z -1 A2(z)= A1(z)+k2B1(z)z -2 = 1+0.78z -1 +0.3z -1 -2 B2(z) =0.3+0.78z + z -1 A3(z)= A2(z)+0.52B2(z)z -3 = 1+0.93z -1 +0.69z -2 +0.5z -2 -3 B3(z) =0.5+0.69z -1 +0.93z + z -1 H3(z)= A3(z)+0.9B3(z)z -4 = 1+1.38z -1 +1.311z -2 +1.337z -3 +0.9z h(n) =1,1.38,1.311,1.337,0.9,0,... . 12.8 Let y(m)= x(2n pm). Then, the backward prediction of x(n p) becomes the forward prediction of y(n). Hence,itslinearprediction error filterisjustthe noisewhiteni ng filterof the corresponding anticausal AR(p) process. 12.9 p x (n + m)= ap(k)x(n -

k) k=1 370 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

e(n)= x(n + m)x (n + m) p = x(n + m)+ ap(k)x(n k) k=1 E[e(n)x * (n l)] =0,l =1,2,. . .,p p . ap(k).xx(k l)= -.xx(l+ m),l =1,2,...,p k=1 The minimum error is E{|e(n)|2} = E[e(n)x * (n + m)] p = .xx(0)+ ap(k).xx(m + k) k=1 Refer to fig 12.9-1. x(n+m) + e(n) z-m-1 forward linear predictor x(n+m) Figure 12.9-1: 12.10 p-1 x (n pm)= bp(k)x(n k) k=0 e(n)= x(n pm)x (n -

pm) p-1 = x(n pm)+ bp(k)x(n k) k=0 E[e(n)x * (n l)] =0,l =0,2,. . .,p 1 p-1 . bp(k).xx(lk)= -.xx(lpm),l =0,2,. . .,p 1 k=0 371 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The minimum error is E{|e(n)|2} = = E[e(n)x * (n pm)] .xx(0)+ p-1 . k=0 bp(k).xx(p+ m k) Refer to fig 12.10-1. x(n) Backward linear predictor x(n-p-m) + e(n) x(n-p-m) z-p-m Figure 12.10-1: 12.11 TheLevinson-Durbinalgorithmfortheforward filter coefficientsis .xx(m)+.bt a m-1 m-1 am(m)= km = f Em * am(k)= am-1(k)+kmam-1(m k), k =1,2,...,m 1; m =1,2,. . .,p * but bm(k)= a (m k),k =0,2,...,m m

b * or am(k) =(m k) m .xx(m)+.t b * m-1 m-1 Therefore, b * (0)= km = Eb m m b * b * (m k)= m-1(m 1k)+kmbm-1(k) m .* (m)+.* bt xx m-1 m-1 Equivalently, bm(0)= k * = m Eb m bm(k)= bm-1(k1)+k * bm* -1(m k) m This is the Levinson-Durbin algorithm for the backward filter. 372 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o

f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

12.12 Let bd m-1 m-1 b= + m 0 bm(m) Then, .b * . . . . ..bdc m-1 m-1 m-1 m-1 m-1 b= = .bt m + 0 bm(m) cm(m) .xx(0) m-1 Hence, (m).b * m-1 ..m-1bm-1 +..m-1dm-1 + bm= c m-1 .bt m-1 + .bt bdm-1 + bm(m).xx(0) = cm(m) m-1

m-1 But..= c m-1bm-1 m-1 . ..= -bm(m).b * m-1dm-1 m-1 Hence, d= -bm(m)..-1 m-1.b * m-1 m-1 .b * Also, ..-1 = a b * m-1 (m).bt b * (m).bt m-1 m-1 Therefore, bmam-1 + bm(m).xx(0) = cmb m-1 solving for bm(m), we obtain (m).bt m-1 m-1 cmbm-1 m-1 bm(m)= b* .xx(0)+.bt a m-1 (m).bt m-1 cmb m-1 m-1 =

f E m-1 we also obtain the recursion * bm(k)= bm-1(k)+bm(m)am-1(m k), k =1,2,...,m 1 12.13 Equations for the forward linear predictor: ..a= c mm m where the elements of care .xx(l+ m),l =1,2,...,p. The solution of ais mm (m).bt cma m-1 am(m)= b* m-1 (0)+.bt a m-1 (m).bt .xxm-1 cma m-1 m-1 = f E m-1 am(k)= am-1(k)+am(m)am* -1(m -

k), k =1,2,...,m 1; m =1,2,...,p where am is the solution to ..a= . mm m b The coefficients for the m-step backward predictor are b= a. mm 373 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

12.14 (a) x (n)= -a1x(n 1)a2x(n 2)a3x(n 3) 14 9 1 But x(n) = x(n 1)+ x(n 2)x(n 3)+w(n) 24 24 24 14 91 E{[x(n)x (n)]2} is minimized by selecting the coefficients as a1 = ,a2 = ,a3 = 24 2424 (b) 3 .xx(m)= ak.xx(m k), m> 0 k=1 p = (m k)+s2 ,m =0

ak.xxw k=1 Since we know the {ak} we can solve for .xx(m),m =0,1,2,3. Then we can obtain .xx(m) for m> 3, by the above recursion. Thus, (c) A3(z) k3 B3(z) A2(z) k2 B2(z) A1(z) k1 12.15 (a) .xx(0) .xx(1) .xx(2) .xx(3) .xx(4) .xx(5) 14 9 1 -2 -3 -1 = 1z z + z 24 24 24 1 = 24 19 14

=4.93 =4.32 =4.2 =3.85 =3.65 =3.46

-2 -3 -1 = z z + z 24 24 24 A3(z)k3B3(z) = 1k32 -2 =10.569z -1 0.351z = -0.351 -1 -2 = -0.3510.569z + z A2(z)k2B2(z) = 1k22 -1 =10.877z = -0.877 4s2 (2z-1)(2z) w ..xx(z)= 9 (3z-1)(3z) = s2 H(z)H(z -1) w

374 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

The minimum-phase system function H(z)is -1 22z H(z)= 33z-1 -1 411 z 2 = 911 z-1 3 (b) The mixed-phase stable system has a system function -1 212z H(z)= 33z-1 -1 212z = 911 z-1 3 12.16 (a) A2(z) = 12rcosz -1 + r 2 z -2 . k2 = r 2 B2(z) = r 2 2rcosz -1 + z -2

A1(z) = A2(z)k2B2(z) 1k2 2 = 12rcos z -1 r2 1+ 2rcos Hence, k1 = 1+r2 (b) Asr . 1,k2 . 1 and k1 .-cos 12.17 (a) a1(1) = -1.25,a2(2)=1.25,a3(3)= -1 -3 Hence, A3(z) =11.25z -1 +1.25z -2 z First, we determine the reflection coefficients. Clearly, k3 = -1, whcih implies that the roots of A3(z)are on the unit circle. We may factor out one root. Thus, -1 A3(z) =(1z -1)(11 -2) z + z 4 -1) = (1z -1)(1az-1)(1a * z v 1+j 63

where a = 8 Hence, the roots of A3(z)are z =1,a, and a* . (b) The autocorrelation function satisfies the equations 3 X. s2 ,m =0 w .xx(m)+ a3(k).xx(m k)= 0, 1 = m = 3 k=1 375 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. .. .. . s2 w .xx(0) .xx(1) .xx(2) .xx(3) 1 .xx(1) .xx(0) .xx(1) .xx(2) .xx(2) .xx(1) .xx(0) .xx(1) -1.25 1.25 0 0 .. .. .. .. = .. .. .xx(3) .xx(2) .xx(1) .xx(0) -10 ff (c) Note that since k3 = -1, the recursion Ef = Em-1(1-|km|2) implies that E= 0. This m 3 f implies that the 4x4 correlation matrix ..xx is singular. Since E0, then s2 =0 3 = w

12.18 .xx(0) =1 .xx(1) = -0.5 .xx(2) =0.625 .xx(3) = -0.6875 Use the Levinson-Durbin algorithm .xx(1) 1 a1(1) = = .xx(0) 2 1 -1 A1(z) =1+ z 2 1 . k1 = 2 3 2 E1 = (1a1(1)).xx(0)= 4 .xx(2)+a1(1).xx(1) 1 a2(2) = = E1 2 1 a2(1) = a1(1)+a2(2)a1(1)= 4 11 -2 -1 Therefore,A2(z) =1+

z z 42 1 . k2 = 2 9 2 E2 = (1a2(2))E1 = 16 .xx(3)+a2(1).xx(2)+a2(2).xx(1) 1 a3(3) = = E2 2 3 a3(2) = a2(2)+a3(3)a2(1)= 8 a3(1) = a2(1)+a3(3)a2(2)=0 31 -2 -3 Therefore,A3(z) =1z + z 82 1 . k3 = 2 27 2 E3 = (1a3(3))E2 = 64

376 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

12.19 (a) 8 ..xx(z)= .xx(m)z -8 -m -1 8 1 1 )-m -m z -m )m z = ( +( 4 4 -8 0 = 1 4z 11 4z + 1 11 4z-1 15 16 = (11 z)(11 z-1)

44 since ..xx(z)= s2H(z)H(z -1), 0.968 H(z)= 11 z-1 4 is the minimum-phase solution. The difference equation is 1 x(n)= x(n 1)+0.968w(n) 4 where w(n)is a white noise sequence with zero mean and unit variance. (b) Ifwe choose 1 H(z)= 11 z 4-1 z = z-1 1 4 -1 4z = 14z-1 then, x(n) =4x(n 1)4 0.968w(n 1)

12.20 .xx(0) =1 .xx(1) =0 2 .xx(2) = -a .xx(3) =0 .xx(1) a1(1) = =0 .xx(0) A1(z) =1 . k1 =0 2 E1 = (1a1(1)).xx(0)=1 .xx(2)+a1(1).xx(1) 2 a2(2) = = a E1 a2(1) = a1(1)+a2(2)a1(1)=0 377 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

2 -2 Therefore,A2(z) =1+az . k2 = a 2 24 E2 = (1a2(2))E1 =1a .xx(3)+a2(1).xx(2)+a2(2).xx(1) a3(3) = =0 E2 2 a3(2) = a2(2)+a3(3)a2(1)= a a3(1) = a2(1)+a3(3)a2(2)=0 2 -2 Therefore,A3(z)= A2(z) =1+az . k3 =0 E3 = E2 =1a 4 12.21 -1 Ap(z)= Ap-1(z)+kpBp-1(z)z where Bp-1(z)is the reverse polynomial of Ap-1(z). For |kp| < 1, we have all the roots inside the unit circle as previously shown. For |kp| =1, Ap(z)is symmetric, which implies that all the roots are on the unit circle. For |kp| > 1, Ap(z)= As(z)+ oBp-1(z)z-1, where As(z) is the symmetric polynomial with all the roots on the unit circle and Bp-1(z)has allthe roots outside the unit circle . Therefore, Ap(z) will have all its roots outside the unit circle. 12.22 1 km V= m

k* 1 m . . . . t* 1 km 10 1 km VJV= m m k* 10 -1 k* 1 mm . . . 1 -km 1 km 1-|km|2 0 == k* -1 k* 10 -(1-|km|2) mm 10 =(1-|km|2) =(1-|km|2)J 0 -1 12.23 (a) m E[fm(n)x(n i)] = E[ am(k)x(n k)x(n i)] k=0 =0, by the orthogonality property 378 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) m * E[gm(n)x(n i)] = a (k)E[x(n m + k)x(n i)] m k=0 m * = am(k).xx(km + i) k=0 = 0, i = 0,1,. . . ,m 1 (c) m E[fm(n)x(n)] = E{fm(n)[fm(n)k=1 am(k)x(n k)]} = = E{|fm(n)|2} Em m-1 E[gm(n)x(n m)] = = = E{gm(n)[gm(n)k=0 bm(k)x(n k)]} E{|gm(n)|2} Em (d) j E[fi(n)fj(n)] = E{fi(n)[x(n)+ k=1 aj(k)x(n k)]} = = E{fi(n)x(n)} Ei = Emax(i,j) where i > j has been assumed (e)

j E[fi(n)fj(n t)] = E{fi(n)[x(n t)+ aj(k)x(n t k)]} k=1 when 0 = t = ij,x(n t1),x(n t2),...,x(n tj)arejust a subset of x(n 1),x(n 2),...,x(n i)Hence, from the orthogonality principle, E[fi(n)fj(n t)]=0 Also, when -1 = t = ij holds, via the same method we have E[fi(n)fj(n t)]=0 (f) j-1 E[gi(n)gj(n t)] = E{gi(n)[x(n t j)+ bj(k)x(n t k)]} k=0 when 0 = t = ij,{x(nt),x(nt1),...,x(ntj)} is a subset of {x(n),...,x(ni+1)}Hence, from the orthogonality principle, E[gi(n)gj(n t)]=0 379

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Also, when 0 = t = ij+1 we obtainthe same result(g) for i = j,E{fi(n + i)fj(n + j)} = E{fi 2(n + i)} = Ei for i 6j, suppose that i >j. Then = j E{fi(n + i)fj(n + j)} = = E{fi(n + i)[x(n + j)+ k=1 aj(k)x(n + jk)]} 0 (h) suppose i > j j-1 . E{gi(n + = E{gi(n k=0 bj(k)x(n k)]} = E[gi(n = Ei (i) for i = j j-1 E{fi(n)gj(n)} = E{fi(n)[x(n j)+ bj(k)x(n k)]} k=0 = E{fi(n)[bj(0)x(n)]} = kjE[fi(n)x(n)] = kjEi for i <j, i E{fi(n)gj(n)} = i)gj(n + j)} + i)[x(n)+ + j+ i)x(n)]

= E{gj(n)[x(n)+ k=1 ai(k)x(n k)]} 0 (j) i-1 E{fi(n)gi(n 1)} = E{fi(n)[x(n 1j)+ k=0 bi(k)x(n 1k)]} = E[fi(n)x(n 1i)] i = E{fi(n)[gi+1(n)k=0 bi+1(k)x(n k)]} = -E[fi(n)bi+1(0)x(n)] = -ki+1Ei (k) i+1 E{gi(n 1)x(n)} = E{gi(n 1)[fi+1(n)k=1 ai+1(k)x(n k)]} = -E[gi(n 1)ai+1(i+ 1)x(n 1i)] = -ki+1Ei i E{fi(n +1)x(n i)} = E{fi(n +1)[fi(n i)ai(k)x(n ik)]} k=1 380 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(l) suppose i>j j-1 E{fi(n)gj(n 1)} = E{fi(n)[x(n 1j)+ bj(k)x(n 1k)]} k=0 =0 Now, let i = j.then i E{fi(n)gj(n 1)} = E{gj(n 1)[x(n)+ ai(k)x(n k)]} k=1 = E{gj(n 1)x(n)} = -kj+1Ej from(d) 12.24 (a)E[fm(n)x *(n i)]=0, 1 = i = m (b)E[gm(n)x *(n i)]=0, 0 = i = m 1 (c)E[fm(n)x *(n)]= E[gm(n)x *(n m)]= Em (d)E[fi(n)fj*(n)]= Emax(i,j) (e) . 1 = t = ij, i>j E[fi(n)fj* (n -

t)]=0, for -1 = t = ij, i<j (f) . 0 = t = ij, i>j * E[gi(n)gj(n t)]=0, for 0 = t = ij+1, i<j (g) . Ei,i = j E[fi(n + i)fj* (n + j)]= 0,i 6= j * (h)E[gi(n + i)gj(n + j)]= Emax(i,j) (i) . k* jEi,i = j * E[fi(n)gi (n)]= 0, i<j * (j)E[fi(n)gi (n 1)] = -ki* +1Ei (k)E[gi(n 1)x *(n)]= -k* Ei i+1 (l) . 0, i<j *

E[fi(n)gj(n 1)] = -k* i = j j+1Ej, 12.25 . 0 .xx(1) .xx(2) .xx(3) . G0 = .xx(0) .xx(1) .xx(2) .xx(3) . 0 .xx(1) .xx(2) .xx(3) . G1 = 0 .xx(0) .xx(1) .xx(2) 381 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

.xx(1) . 1 k1 . k1 = V1 = .xx(0) k1 * 1 . 00 .xx(2)+k1.xx(1) .xx(3)+k1.xx(2) . V1G1 = 0 .xx(0)+k1*.xx(1) .xx(1)+k1*.xx(2) .xx(2)+k1*.xx(3) . 00 .xx(2)+k1.xx(1) .xx(3)+k1.xx(2) . G2 = 00 .xx(0)+k1*.xx(1) .xx(1)+k1*.xx(2) 2 (1) xx(2)xx(1) (2)+k1(0) xx xxxx Therefore, k2 = = 2 (1)xx(1) xx(0)+k* (0) 1xxxx Let, 1 k2 V= 2 k* 1 2 00 0 A V2G2 = 00 k2[.xx(2)+k1.xx(1)]+.xx(0)+k1.xx(1) B where A = .xx(3)+k1.xx(2)+k1k2.xx(2)+k2.xx(1), and B = k2.xx(3)+k1k2.xx(2)+k1.xx(2)+.xx(1) . 000 .xx(3)+k1(1+k2).xx(2)+k2.xx(1) .

and therefore, G3 = 000 k2.xx(3)+k1(1+k2).xx(2)+.xx(1) and [.2 (1).2 (0)].xx(3)+C xxxx k3 = [.xx(0).xx(2).2 (1)].xx(3).xx(1).2 (2)+D xxxx where C =2.xx(0).xx(1).xx(2).xx(1).2 (2).3 (1)and xxxx D = (0).xx(1).xx(2)+.3 (1).xx(1).2 (0) .xxxxxx This is the same result obtained from the Levinson Algorithm. 12.26 The results of section 11.1 apply directly to this problem. We may express ..xx( f)as s2 ..xx(f)= w|H(f)|2 where H(f)is a filter with transfer function 8 -m] H(z)= exp[ v(m)z m=1 Theprediction error filterwhitens theinputprocess,sothattheoutputprocessiswhitew ith spectral density s2 = exp[v(0)]. Therefore, the minimum MSE is w .

Ef = s2 dw 8 w . = s2 dw w 382 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

=2ps2 w 2pev(0) = 1 2 But v(0) = ln..xx(f)df 1 2 1 2 Therefore, Ef =2pexp[ ln..xx(f)df] 8 12.27 1 2 12.28 (a) ..xx(z) . G(z) Since d(n) .dx(k) Therefore, ..dx(z) . ..dx(z) G(z-1) + Hopt(z) hopt(n) the output is y(n) MMSE8

G0 = = = = = = = = = = = = = = = = = = = = = 0 1 s2 w (1az-1)(1az) 1 1az-1 x(n + m), we have E{d(n)x(n k)} E{x(n + m)x(n k)} .xx(m + k) z M..xx(z) . zms2 (1az) .

w (1az-1)(1az) + . s2 m . z w 1az-1 + m a s2 w 1az-1 m 1 -1) as2 (1az w s2 1az-1 w m a a md(n) hopt(n)* x(n) m ax(m) 8

.xx(0)hopt(k).dx(k) k=0 .xx(0)a m.dx(0) .xx(0)a m.xx(m) ms2 s2 a w mw a 1a2 1a2 2m 1a s2 w 1a2 11 1 . 2 8 64 11 1 2 8 64 383 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. 11 1 . 1 0 2 8 64 G1 = . k1 = 11 01 2 28 . 1 1 2 V= 1 1 1 2 . 1 3 00 8 64 V1G1 = 37 15 0 4 16 128 . 1 3 . 2

00 8 64 G2 = 37 . k2 = 00 3 4 16 . 2 1 3 V= 22 1 3 . 47 000 192 V2G2 = 2 13 00 3 32 . 47 . 47 000 = 192 . k3 = G32 000 128 3

(b)Refer to fig 12.28-1 12.29 8 -j2fk ..xx(f)= .xx(k)e k=-8 m -j2fp Am(f)= am(p)e p=0 n * j2fq A * (f)= a (q)e nn a=0 1 2 1 2 * (f)A * n(f)df = .xx(k)am(p)an(q) e -j2f(k+p+q)df ..xx(f)Am 1 2 1 2 kpq 8 mn * = .xx(k)am(p)an(q)d(qpk) k=-8 p=0 q=0 mn

* = .xx(qp)am(p)a (q) n p=0 q=0 * = E[x(l+ q)x * (l+ p)]am(p)a (q) n pq 384 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + +z -1 + z -1 + + z -1 w(n) x(n) -47/128 -2/3 1/247/128 2/3 -1/2 + + +z -1 + z -1 + + z -1 w(n) x(n) -47/128 -2/3 1/247/128 2/3 -1/2 Figure 12.28-1: . mn . * = Eam(p)x * (l+ p) a (q)x(l+ p) n p=0 q=0 = E{fm(l+ m)f * (l+ n)} n = Emdmn wherethelaststepfollowsfrom prob.12.24 property(g) 12.30 A1(z) B1(z) A2(z) B2(z) A3(z) B3(z) A4(z)

H(z) -1 = 1+0.6z -1 =0.6+z -1 = A1(z)+k2B1(z)z -2 = 1+0.78z -1 +0.3z -1 -2 =0.3+0.78z + z -1 = A2(z)+k3B2(z)z -3 = 1+0.93z -1 +0.69z -2 +0.5z -2 -3 =0.5+0.69z -1 +0.93z + z -1 = A3(z)+k4B3(z)z = 1+1.38z -1 +1.311z -2 +1.337z 1 = A4(z) 385 -4 -3 +0.9z (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

12.31 A2(z) = 1+ 0.1z -1 0.72z -2 k2 = -0.72 B2(z) = -0.72+ 0.1z -1 + z -2 A2(z)k2B2(z) A1(z) = 1k2 2 = 1+ 0.357z -1 k1 = 0.357 B1(z) = 0.357+z -1 A0(z) = B0(z)= 1 C2(z) = 0B0(z)+1B1(z)+2B3(z) = 0 + 1(0.357+z -1)+2(-0.72+ 0.1z -1 + z -2) = 10.8z -1 + 0.15z -2 Hence, 0 = 1.399 1 = -0.815 2 = 0.15 Refer to fig 12.31-1 + + +z -1 + z -1 + + -0.72 0.357 0.15 -0.815 1.399 0.72 -0.357 output input Figure 12.31-1: 12.32 Refer to fig 12.32-1 ht(n)mininizes E[e2(n)](wienerfilter) length M =2(a) 386 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

w(n) + +h+ (n) s(n) x(n) y(n) + e(n) d(n) FIR Figure 12.32-1: s2 ..ss(w)= v|H(w)|2 0.49 = |10.8e-jw|2 0.49 ..ss(z)= (10.8z-1)(10.8z) We can either formally invert this z-transform, or use the following idea: The i nverse ztransform of 12.1 will have the form .ss(m)= .ss(0)(0.8)|m| From the AR model for s(n)it is easy to show .ss(0) =0.8.ss(1)+.sv(0) =0.8.ss(1)+d2 and .ss(1) =0.8.ss(0)+.sv(1) =0.8.ss(0) 49 solve for .ss(0) = 36 49 4)|m| so .ss(m)= ( 365Now .ss(m)= E[x(n)x(n m)] = E{[s(n)+w(n)][s(n m)+w(n m)]}

=(m)+s2 d(m) .ssw 49 4)|m| = ( + d(m) 365 (b) d(n)= s(n) .dx(l)= .sx(l)= E[s(n)x(n l)] = E{s(n)[s(n l)+w(n l)]} = .ss(l) So the normal equations are . 49 49 4 . . 49 1+ ht(0) 36 36 5 36 = 49 4 49 49 4 ht(1) 1+ 36 5 36 36 5 387 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

ht(0)=0.462,ht(1)=0.248 49 49 49 4 (c). =MMSE2 = 0.462 x 0.248 x x =0.462 36 36 36 5 12.33 ..dx(z) =..ss(z) 0.49 = (10.8z-1)(10.8z) ..xx(z) =..ss(z)+1 1.78(10.45z-1)(10.45z) = (10.8z-1)(10.8z) (10.45z-1) G(z) = (10.8z-1) . ..dx(z). = . 0.49 . G(z-1) + (10.8z-1)(10.45z) + . 0.766 0.345z . = 10.8z-1 + 1-

0.45z + 0.766 = 10.8z-1 H+ c (z) = 1 1.78 10.8z-1 10.45z-1 0.766 10.8z-1 0.43 = 10.45z-1 h+ c (n) = 0.43(0.45)n u(n) .+ c = MMSE8 = 1 2pj . c [..ss(z)Hc(z)..ss(z -1)]z -1dz 1 . 0.28 = dz 2pj c (z 0.45)(10.8z) = 0.438 12.34 Using quantities in prob. 12-33, ..dx(z) H+ (z)= nc ..xx(z) 0.275 = (10.45z-1)(10.45z) .+

=MMSEnc = 1 [..dd(z)H+ (z)..dx(z -1)]z -1dz nc nc 2pj c 10.275 = dz 2pj (z 0.45)(10.45z-1) c =0.345 388 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

12.35 .ss(m)=(0.6)|m| . .. ... 20.60.36 h+(0) 1 . 0.620.6 .. h+(1) . = . 0.6 . 0.36 0.62 h+(2) 0.36 h+(0)=0.455,h+(1)=0.15,h+(2)=0.055 .3 =MMSE3 =10.4550.15 x0.60.055 x0.36 =0.435 Increasing the length of the filter decreases the MMSE. 12.36 . .. ... .xx(0) .xx(1) .xx(2) 1 1 . .xx(1) .xx(0) .xx(1) .. -1 . = . 0 . s2 w .xx(2) .xx(1) .xx(0) 0.60 . .. ... 1 -10.6 .xx(0) 1 -11.60(1) 0 . s2 . . .. .xx. = . w 0.6 -11 .xx(2) 0 .xx(0)=2.5641,.xx(1)=1.6026,.xx(2)=0.064

For m = 3,.xx(m)= .xx(m 1)0.6.xx(m 2) For m< 0,.xx(m)= .xx(-m) 12.37 s2 v ..ss(z)= (1+Pp (k)z-k)(1+Pp (k)zk) k=1 apk=1 ap . Let ap(0) =1 s2 v ..ss(z)= (Pp (k)zk) (k)z-k)(Pp k=0 apk=0 ap ..ss(z) =..ss(z)+s2 w s2 + s2 (.)(.) vw = (.)(.) 389 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x(n)is ARMA(p,p). Suppose (Pp bp(k)z-k)(Pp bp(k)zk) k=0 k=0 ..xx(z)= (Pp (k)z-k)(Pp (k)zk) k=0 apk=0 ap Comparing parameters of the two numerators pp s2 + s2 . s2 . b2 2 a (k) =(k) v wpnp k=0 k=0 p-qp-q s2 . s2 . ap(k)ap(k+ q)= bp(k)bp(k+ q) q =1,2,...,p wn k=0 k=0 There are p+1 equations in p+1 unknown parameters s2,bp(1),...,bp(p). Note that bp(0)=1. n 390 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 13 13.1 N M-1 #2 E = y(n)+w(n)h(k)x(n k) n=0 k=0 By carrying out the minimization we obtain the set of linear equations: M-1 h(k)rxx(lk)= ryx(l)+rwx(l),l =0, ,M 1 k=0 where, N rwx(l)= w(n)x(n l) n=0 13.2 If we assume the presence of a near-end echo only, the received signal is 8 rA(t)= AsA(t d1)+w(t)= Aa(k)p(t d1 kTs)+w(t) k=0 The receiver filter eliminates the noise outside the frequency band occupied by the signal and after sampling at the symbol rate we obtain, 8

r(n)= Aa(k)p(nTs d1 kTs)+w(nTs). k=0 If we assume that the delay d1 is a multiple of the symbol time interval, that i s, d1 = DTs, then, r(n)= Aa(n D)+w(n) The LS criterion minimizes E = 8 n=0 M-1 r(n)h(k)a(n k) k=0 2 . Theequationsforthe coefficientsoftheadaptive echo cancelerare M-1 h(k)raa(lk)= rra(l) l =0, ,M 1 k=0 391 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

where, raa(lk)= a(n k)a(n l). n rra(l)=[Aa(n D)+w(n)]a(n l)= Araa(lD)+rwa(l). n 13.3 Assumethatthesampleautocorrelationand crosscorrelationaregivenbythe unbiased est imates: N-1 N-1 1 . 1 . rvv(k)= v(n)v(n k) ryv(k)= y(n)v(n k) NN n=0 n=0 Then, N-1. 8 !. 8 . 1 X. . rvv(k)= w2(n l)h(l)+w3(n) w2(n kp)h(p)+w3(n k)

N n=0 l=0 p=0 N-1. 88 1 . X. = w2(n l)w2(n kp)h(l)h(p)+ N n=0 l=0 p=0 88 + w2(n l)w3(n k)h(l)+ h(p)w2(n kp)w3(n)+w3(n)w3(n k) l=0 p=0 Since E[w2(n l)w3(n k)]=0, we obtain N-1. 88 . 1 . X. E[rvv(k)]= E[w2(n l)w2(n kp)]h(l)h(p)+E[w3(n)w3(n k)] N n=0 l=0 p=0 or 88 E[rvv(k)]= h(l)h(p).w2w2(k+ p-

l)+.w3w3(k) l=0 p=0 N-1 1 . ryv(k)= y(n)v(n k) N n=0 N-1 . 8 . 1 X. = (x(n)+w1(n)+w2(n)) w2(n kl)h(l)+w3(n k) N n=0 l=0 N-1 8 1 X. = (x(n)+w1(n)+w2(n))w2(n kl)h(l)+ N n=0 l=0 N-1 1 . + (x(n)w3(n k)+w1(n)w3(n k)+w2(n)w3(n -

k)) N n=0 8 =. E[ryv(k)] =[.xw2(k+ l)+.w2w2(k+ l)]h(l)+.xw3(k) l=0 Further simplifications are obtained if w1,w2,w3 are white and x is uncorrelated with w2. 392 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

13.4 We need to prove that H [1a (n)]Vm+1(n) =[b+ (n 1) 1]Q * = K * mmm+1(n)m+1(n) . x *(l) . nn . X. x *(l1) . n-ln-l Vm+1(n)= wx(lm 1)X * = wx(lm 1) m+1(n).. .. .. l=0 l=0 . .. x *(lm) .Pnn-l.. . x(lm 1)x *(l) v l=0 w = . . = . . Vm(n 1) Vm(n 1)

Thus, .. v H [1a (n)]Vm+1(n) =[1 QH(n)R-1(n 1)] . . mmm Vm(n 1) = v QH(n)R-1(n 1)Vm(n 1) = v + QH(n)bm(n 1) mm m = v + bt (n 1)Q * (n) mm But, n . Qm(n) . n-lX(l)X * Qm+1(n)= w m+1(l1) = . . * l=0 v Hence, . Q* (n) . m H

[1a (n)]Vm+1(n)=[bt (n 1) 1] . . =[bt (n 1) 1]Q * mmmm+1(n) v From the definition of Km+1(n)in(13.3.29) we obtain H [1a =[bt (n 1) 1]Q * = K * m(n)]Vm+1(n)mm+1(n)m+1(n) 13.5 We need to prove that * am(n)gm(n)em+1(n) .m(n)= .m(n 1)Eb (n) m By definition dm(n) .m(n)= Eb (n) m Use the relations: * (n)= wdm(n 1)+amg (n)em(n) dmm dm(n 1)gm(n) em+1(n)= em(n)Eb (n -

1) m Eb (n)= wEb (n 1)+am|gm(n)|2 mm m(n-1)gm(n) * wdm(n 1)+amg (n) em+1(n)+ mEb (n-1) .m(n)= m Eb (n) m 393 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

* (n 1)(wEb (n 1)+(n)|2)(n)g dmmam(n)|gmamm(n)em+1(n) = -Eb (n)Eb (n 1) Eb (n) mmm * dm(n 1) am(n)gm(n)em+1(n) = -Eb (n 1) Eb (n) mm * am(n)g (n)em+1(n) m = .m(n 1)Eb (n) m 13.6 * um(n) wum(n 1)+2fm-1(n)gm-1(n 1) Km(n)== =. vm(n) vm(n) **

vm(n)Km(n)= wum(n 1)+fm-1(n)gm-1(n 1)+fm-1(n)gm-1(n 1) * = wum(n 1)+[fm(n)+Km(n)gm-1(n 1)]gm-1(n 1) * (n)f * +[g (n)+Kmm-1(n)]fm-1(n) m ** = wum(n 1)+fm(n)gm-1(n 1)+g (n)fm-1(n) m 2 + Km(n){|gm-1(n 1)| + |f2 m-1(n 1)|} 2 Km(n)[vmm-1(n 1)|-|f2 (n)-|g m-1(n 1)|] ** = wum(n 1)+fm(n)gm-1(n 1)+g (n)fm-1(n) m ** =. Km(n)wvm(n 1) = wum(n 1)+fm(n)gm-1(n 1)+g (n)fm-1(n)

m ** (n)gm-1(n 1)+g fmm(n)fm-1(n) Km(n)= Km(n 1)+ wvm(n 1) 13.7 We will derive the FAEST algorithm in Table 13.7 line by line. The alternative K alman gain is defined as Km(n)= 1 Pm(n 1)X * (n) m w From(13.2.74) Pm(n 1)X* (n) K m(n) m Km(n)= = w + Xt (n)Pm(n 1)X* (n) 1+ 1 Xt (n)Pm(n 1)X* (n) mmmm w

= Km(n)am(n) (see13.3.57) Define am(n)=1/am(n). Then, (n)=1+ 1 Xt (n)Pm(n 1)X * (n)=1+Xt (n) (n). ammmmKm

w FAEST-line 1: t fm-1(n)= x(n)+am-1(n 1)Xm-1(n 1) FAEST-line 2: fm-1(n) f m-1(n,n)= fm-1(n)am-1(n)= . a m-1(n 1) FAEST-line3: From(13.3.50) am(n)= am(n 1)Km(n 1)fm(n)=. am-1(n)= am-1(n 1)K m-1(n)am-1(n 1)fm-1(n) = am-1(n 1)K m-1(n)f m-1(n,n) 394 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

FAEST-line4: From(13.3.83) ff E(n)= wE(n 1)+am-1(n 1)fm-1(n)f * m-1m-1m-1(n) But am-1(n 1)f* = f * m-1(n)m-1(n,n), thus, ff E(n)= wEm-1(n,n) (n 1)+fm-1(n)f * m-1m-1 FAEST-line 5: R-1 Km(n)= 1 R-1(n 1)X * (n) and K m-1(n 1) = 1 (n 2)X * m-1(n 1) mmm-1 ww Usethe partition(13.3.32) towrite 00 11 R-1 H (n 1) = +[1am-1(n 1)]. m R-1 f

0(n 2) am-1(n 1) m-1E(n 1) m-1 Thus, 100 11 H

Km(n)= + [1am-1(n 1)] R-1 f 0(n 2) am-1(n 1) w m-1wE(n 1) m-1 x *(n) X* m-1(n 1) . f* 0 m-1(n)1 =+ f Km-1(n 1) am-1(n 1) wE(n 1) m-1 FAEST-line 6: We need to find the update formula for the step K m+1(n +1) -. K

m(n +1). Km(n)= 1 Pm(n 1)X * (n). m w Using partition(13.3.27) we obtain Pm-1(n 1)0 1 bm-1(n 1) [bH Pm(n 1) = + m-1(n 1) 1] 00 Eb (n 1) 1 m-1 Thus, . . X* 1 Pm-1(n 1) 0 m-1(n)1

Km(n)= + w 00 x *(n m +1) wEb (n 1) m-1 X* bm-1(n 1) m-1(n) [bH

m-1(n 1) 1] 1 x *(n m +1) . * K m-1(n 1) gm-1 bm-1(n 1) =+ 0 wEb (n 1) 1 m-1 Write

Cm-1(n)

Km(n)= c mm(n) We identify * gm-1(n) * c mm(n)= =gm-1(n)= wEb cmm . m-1(n 1) (n). wEb (n 1) m-1 395 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

FAEST-line 7: Using the partition of K m(n)in step-6 we obtain

Cm-1(n)= Km-1(n)+ cmm(n)bm-1(n 1) =. Km-1(n)= Cm-1(n)c mm(n)bm-1(n 1) FAEST-line8: From(13.3.91) f* m-1(n,n)fm-1(n) am(n)= am-1(n 1) 1f E(n) m-1 But, am(n)=1/a m(n)and fm-1(n)= f 1). Thus, Ef (n) m-1 a m(n)= a m-1(n 1) f E(n)am-1(n 1)|fm-1(n)|2 m-1 f wE(n 1)+|fm-1(n)|2/a m-1(n 1) m-1 = a m-1(n 1) f wE(n) m-1(n,n) am-1(n -

m-1 1 |fm-1(n)|2 = a m-1(n 1) 1+ f a m-1(n 1)wE(n) m-1|fm-1(n)|2 = a m-1(n 1)+ wEf (n) m-1FAEST-line 9: am(n)=1+ m(n)Xm(n). If we use the partition of step-6 then Kt * gm-1(n) Kt [bt wEb (n 1) a m(n) = 1+[ m-1(n)0]Xm(n)+ m-1(n 1) 1]Xm(n) m-1= a m-1(n)+ cmm(n)gm-1(n)=. a m-1(n)= a m(n)c mm(n)gm-1(n) FAEST-line10: From(13.3.61) gm-1(n) g m-1(n,n)= gm-1(n)am-1(n)= a m-1(n) FAEST-line11: From(13.3.84) * Eb (n)= wEb (n 1)+am(n)g (n)gm(n). mmm But, gm(n)am(n)= g m(n,n), so that, ** Eb (n)= wEb (n 1)+g (n) gm(n,n)= wEb (n 1)+gm(n) g (n,n)

mmmmm FAEST-line 12: The time-update of bm(n)is given by(13.3.51) bm(n)= bm(n 1)Km(n)gm(n) But, Km(n)= Km(n)am(n)and am(n)gm(n)= g m(n,n), so that bm(n)= bm(n 1)K m(n) gm(n,n) FAEST-line 13: By definition eM(n)= d(n)ht (n 1)Xm(n) m FAEST-line14,15: From(13.2.76) hm(n)= hm(n 1)+Km(n)em(n)= hm(n 1)+ K m(n)am(n)em(n) = hm(n 1)+ K m(n) em(n,n) where, em(n) e m(n,n)= am(n)em(n)= a m(n) 396 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

13.8 hm(n +1) = whm(n)+.e(n)X * (n) m where e(n)= d(n)hT (n)Xm(n). m Thus, hm(n +1) = whm(n)+.(d(n)hT (n)Xm(n))X * (n) mm = whm(n)+.d(n)X * (n)-.X * (n)Xt (n)hm(n) mmm =. E[hm(n +1)] =(wI -.Rm)E[hm(n)]+.rm where, Rm = E[X * (n)Xt (n)],rm = E[d(n)X * (n)]. mmm Since Rm isHermitian,it assumesthe decomposition Rm = UUH,whereisa diagonal matrix with elements .k,0 = k = m 1, the eigenvalues of Rm, and U is a normalized modal matrix such that UUH = I. Thus, E[hm(n +1)] = U[wI -.]UHE[hm(n)]+.rm. Premultiplying the above by UH we obtain 00 0 h (n +1) =[wI -.]h (n)+.r mmm 00 UH where h (n +1) = UHE[hm(n +1)], r= rm. The values of . that ensure convergence of

mm the mean of the coefficient vector should satisfy |w -..k| < 1,k =0, ,m 1 or 1w 1+w < . < ,k =0, ,m 1 .k .k or 1w 1+w < . <. .min .max 13.9 e(n)= |e(n)|2 + ckhM(n)k2 d * (n)hH M(n) M(n)hM(n) = d(n)XM T (n)hM(n)M(n)X * + chH . hH . = |d(n)|2 2Re M(n)X * M(n) d(n)+hH M(n)X * M(n)XM T (n)hM(n)+chH M(n)hM(n) The complex gradient vector is .e(n)/.hH M: .e(n) M(n)XT .hH = -XM* (n)d(n)+X * M(n)hM(n)+chM(n)

M -X * . = M(n)d(n)+XM T (n)hM(n)+ chM(n) = -e(n)XM* (n)+chM(n) 397 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Then,inthesteepest-descentmethod,weupdatethe coeficientvectorasfollows: .e(n) hM(n +1) = hM(n).hH M = hM(n)+[e(n)X * M(n)chM(n)] = (1c)hM(n)+e(n)X * M(n) 13.10 The normalized LMS algorithm is given as: X * hM(n +1) = hM(n)+ M(n)e(n) kX(n)k2DEfine the error vector e(n)as e(n)= hopt(n)hM(n) Also, define the mean square derivation of the error vector as J(n)= Eke(n)k2. Then, X * kX(n)k2 J(n +1) = E ke(n)M(n)e(n)k2 ReeH(n)X * (n)e(n). |e(n)|2 = J(n)2E +2E kX(n)k2 kX(n)k2 Hence, . |e(n)|2 ReeH(n)X * (n)e(n) 2 kX(n)k2 kX(n)k2

We observe that the mean square derivation decreases exponentially with an incre ase in n, provided that . Re["H(n)X * (n)e(n)]. J(n +1) J(n)= 2E E kX(n)k2 0 < <

. |e(n)|2 E kX(n)k2 Approximation: . |e(n)|2 E|e(n)|2. E kX(n)k2 E[kX(n)k2] and . ReeH(n)X * (n)e(n)EReeH(n)X * (n)e(n) E

kX(n)k2 E[kX(n)k2] With the approximations, we obtain ReeH(n)X * (n)e(n) 0 < <

E[|e(n)|2] 398 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

13.11 We can reduce the number of computations needed by m 1 multiplications if we avoid the update of the Kalman gain Km-1(n)= xCm-1(n)ybm-1(n 1). If we use the alternative Kalman gain this step takes the form Km-1(n)= Cm-1(n)ybm-1(n 1).

As in the a-priori case, the update of the alternative Kalman gain vector Km(n), is carried out in two steps, step-up

-. Km+1(n +1)

Km(n) step-down

-. Km(n +1) using the following Levinson-type recursions: . f* 0 m-1 1

Km(n)= + (step5ofprob.138) f Km-1(n 1) am-1(n 1)

wE(n 1) m-1 and . * . K m-1(n) gm-1 bm-1(n 1) Km(n)= + (step6ofprob.138) 0 wEb (n 1) 1 m-1 With K m(n)we associate the scalar (n)= 1 =1+X * (n) (n). ammKm am(n) This parameteris updatedas(see prob.13.8) |fm-1(n)|2 a m(n)= a m-1(n 1)+ f wE(n) m-1 a m-1(n 1) = a m(n)gm-1(n) cmm(n). FASTRLS algorithm: VersionA(a-posteriori version) t fm-1(n)= x(n)+am-1(n 1)Xm-1(n 1) gm-1(n)= x(n am(n)

M +1)+ bm t -1(n 1)Xm-1(n 1) fm-1(n) am-1(n)= am-1(n 1)K m-1(n 1) a m-1(n 1) t fm-1(n,n)= x(n)+am-1(n 1)Xm-1(n 1) ff E(n)= wEm-1(n,n) (n 1)+fm-1(n)f * m-1m-1 f* Cm-1(n)0 m-1(n)1

Km(n)= =+ c mm(n) K m-1(n 1) f am-1(n 1) wE(n 1) m-1

Km-1(n)= Cm-1(n)bm-1(n 1) cmm(n) |fm-1(n)|2 a m(n)= a m-1(n 1)+

f wE(n 1) m-1 a m-1(n)= a m(n)gm-1(n) cmm(n) gm-1(n) bm-1(n)= bm-1(n 1)K m-1(n 1) a m-1(n) 399 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

ht d(n) =(n 1)Xm(n) m em(n)= d(n)d (n)

Km(n)em(n) hm(n)= hm(n 1)+ a m(n) Initialization: f

am-1(-1) = bm-1(-1) =0,Km-1(-1) =0,hm-1(-1) =0,E(-1) = E > 0. m-1 fm-1(n) |fm-1(n)|2 In this version we need 5 extra multiplications for the calculation of , , f m-1 em(n) a m-1(n-1)wE(n-1) gm-1(n) (n), gm-1(n) and we save m multiplicationsfromthe estimationof K m-1(n). cmma m-1(n-1), a m(n) FASTRLS algorithm: VersionB(a-posteriori version) t fm-1(n)= x(n)+am-1(n 1)Xm-1(n 1) gm-1(n)= x(n M +1)+ bm t -1(n 1)Xm-1(n) fm-1(n)

am-1(n)= am-1(n 1)K m-1(n 1) a m-1(n 1) fm-1(n) f m-1(n,n)= a m-1(n 1) |fm-1(n)|2 ff E(n)= wE(n 1)+ m-1m-1 a m-1(n 1) f* Cm-1(n)0 m-1(n)1

Km(n)= =+ f c mm(n) Km-1(n 1) am-1(n 1) wE(n 1) m-1

Km-1(n)= Cm-1(n)bm-1(n 1) cmm(n) |fm-1(n)|2 a m(n)= a m-1(n 1)+ f wE(n -

1) m-1 a m-1(n)= a m(n)gm-1(n) cmm(n) gm-1(n) bm-1(n)= bm-1(n 1)K m-1(n 1) a m-1(n) d (n)= ht (n 1)Xm(n) m em(n)= d(n)d (n)

Km(n)em(n) hm(n)= hm(n 1)+ a m(n) Initialization: f

am-1(-1) = bm-1(-1) =0,Km-1(-1) =0,hm-1(-1) =0,E(-1) = E > 0,a m-1(-1) =1. m-1 fm-1(n) gm-1(n) em(n) In this version we need 3 extra multiplications for the calculation of , , a m-1(n-1)a m-1(n-1)a m(n) and we save m multiplications from the estimation of K m-1(n). 13.12 .. M-1 !2. E = E. gh(n)x(n) .

n=0 400 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

". M-1 !. .E . =0 =. E 2 gh(n)x(n) x(k) =0,k =0, ,M 1. .h(k) n=0 Thus, "M-1 . E[gx(k)]= Eh(n)x(n)x(k) ,k =0, ,M 1. n=0 E[gx(k)] = E[g(gv(k)+w(k))]= E[g 2]v(k)+E[gw(k)] = Gv(k) (ifg,w(k)areuncorrelated) "M-1 . M-1 Eh(n)x(n)x(k)= h(n)E[x(n)x(k)] n=0 n=0 M-1 = h(n)E[(gv(n)+w(n))(gv(k)+w(k))] n=0 M-1 2 = h(n)E[gv(n)v(k)+gv(n)w(k)+gv(k)w(n)+w(n)w(k)] n=0 M-1 = Gh(n)v(k)v(n)+s2 h(k) w n=0 Hence,

M-1 Gv(k)= Gv(k) h(n)v(n)+s2 h(k) w n=0 or T (Gvv + s2 I)h = Gv w where v =[v(0), ,v(M 1)]T , h =[h(0), ,h(M 1)]T . 13.13 Let M-1 M-1 -kj2n/M)-j2nk/M H(z)= hkz and Hn = H(z = e= hke. k=0 k=0 The sequence {hk} is related to the sequence {Hn} by the inverse discrete Fourier transform M-1 1 . j2n/M Hne,k =0, ,M 1. hk = M n=0 When hk, given above is substituted in the expression for H(z)the double sum tha t results can be simplified to yield

M-1 1z-M . Hk H(z)= . 1ej2k/Mz-1 M k=0 The filter structure is shown in Fig. 13.13-1. 1. Let yk(n)be the output at time t = nT of the filter with transfer function -M 1z1 . 1ej2k/Mz-1 M 401 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 - z-1 1 - z-M 1 - z-1 H0 y0 X yM-1 HM-1 H1 y1 d(n) ^ 1 1 1 M 1 - z-1 ej2 p X X + + + d(n) e(n) (M-1) e j2 p/M /M X yM-1 HM-1 H1 y1 d(n) ^ 1 1 1 M 1 - z-1 ej2 p X X + + + d(n) e(n) (M-1) e j2 p/M /M Figure 13.13-1: Then the response of the recursive filter at t = nT is M-1

d(n) = Hk(n)yk(n). k=0

= nT. If e(n)= dd (n)then, an algorithm where {Hk(n)} are the filter coefficients at t for adjusting the coefficients Hk(n)is given by Hk(n +1) = Hk(n)+.e(n)yk(n)k =0, ,M 1. 2. The cascade of the comb filter 1-z -M with each of the single-pole filter forms a system M with frequency response j2f/M 1e Hk(f)= . M(1ej2(k/M-f)) Thus, e j2Mf e-j2Mf e-j2(k/M-f) ej2(k/M-f) -j2Mf 1 e |Hk(f)| = ej2(k/M-f) M 1 = 1

M sin(pMf) sin(p(k/M f)) . 2jsin(pMf) -2jsin(p(k/M f)) = M We observe that |Hk(f)| = 0 at the frequencies f = n/M, n 6= k and |Hk(f)| =1 at f = k/M. Thus, the kth system has a resonant frequency at f = k/M, and it is zero at the resonant frequencies of all the other systems. This means that if the desired signal is M-1 2pk , d(n)= Ak cos(.kn),.k = M k=0 the coefficient of each single-pole filter can be adjusted independently without any interaction from the other filters. 402 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

13.14 .J =2h(n)40 .h(n) Thus, h(n +1) = h(n)-.h(n)+20. = h(n)(1-.)+20.. 1. For an overdamped system, |1.| < 1=. 0 < . < 2. 2. Fig. 13.14-1 contains a plot of J(n)vs. n. The step . was setto0.5 and the initial value of h was set to 0. In Fig. 13.14-2 we have plotted J(h(n))vs. h(n). As it is obs erved from the figures the minimum value of J which is -372, is reached within 5 iterations of the algorithm. 50 J(n) -150 -100 -50 0 -350 -300 -250 -200 -400 0 5 10 15 20 25 30 35 40 45 50 n Figure 13.14-1: 13.15 Normal Equations: M-1 . a(k)rvv(lk)= ryv(l) l = 0,1, ,M 1 k=0 403 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho

ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

50 J(h(n)) 0 -50 -100 -150 -200 -250 -300 -350 -400 0 2 4 6 8 10 12 14 16 18 20 h(n) Figure 13.14-2: rvv(lk)= rw3w3(lk)+rv2v2(lk) Power spectral density of v2(n): ..v2v2(f)= s2 w|H(f)|2 = s2 w 1 = s2 w 0.75 1.250.75 cos(2pf) . |10.5e-j2f|2 Thus, s2 w rv2v2(m)= (0.5)|m|. 0.75 Hence, s2

w s2 rvv(lk)= wd(lk)+ (0.5)|l-k|. 0.75Assuming that x(n),w1(n),w2(n),w3(n)are mutually uncorrelated, it follows th at 8 E[y(n)v(n l)]= E[w2v2(n l)]= Ew2 h(k)w2(n lk) , k=0 where h(k)=0.5k . Thus, 88 E[y(n)v(n l)]= h(k)E[w2(n)w2(n lk)]= h(k)s2 d(l+ k)= s2 d(l). ww k=0 k=0 404 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

+ + e(n)-w (n)2 A(z) w (n) v (n)2 3 1 1-0.5 z-1 + e(n)-w (n)2 A(z) w (n) v (n)2 3 1 1-0.5 z-1 y(n)=x(n) + 12

v(n)

+ v(n)

w (n) + w (n)

Figure 13.15-1: The normal equations take the form . .

0.252 0.52 w s2 w + w w w . .. . 0.75 0.75 0.75 s2 w a(0) ... .

. ... . 0.52 2 0.75 0 w .75 w s2 w a(1) 0 . = . . + 0.75 2 w s2 + w 0.75 a(2) 0 0.252 0.52 w w 0.75 0.75 15 4 1 =. a(0)= ,a(1)= ,a(2)= -

. 323232 13.16 e(n)= a1x(n 1)a2x(n 2) E = E[e .E x(n)2(n)] =.

= E[(x(n)a1x(n 1)a2x(n 2))x(n 1)] =0 .a1 .E = E[(x(n)a1x(n 1)a2x(n 2))x(n 2)] =0 .a2 =. E[x(n)x(n 1)]a1E[x(n 1)x(n 1)]a2E[x(n 2)x(n 1)] =0 E[x(n)x(n 2)]a1E[x(n 1)x(n 2)]a2E[x(n 2)x(n 2)] =0 But, E[x(n)x(n 1)] = E[x(n -

2)x(n 1)] = a E[x(n 1)x(n 1)] = E[x(n 2)x(n 2)] =1 2 E[x(n)x(n 2)] = a Thus, we obtain the system . 1 aa1 a = 2 a 1 a2 a with solution a1 = a, a2 =0. 405 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

13.17 The optimum linear predictor in Prob. 13.16 is a first order filter with transfe r function -1 A(z)=1az . Thus, the corresponding lattice has one stage with the forward and backward erro rs given by f(n)= f0(n)+Kb0(n 1) b(n)= b0(n 1)+Kf0(n) Since f0(n)= b0(n)= x(n), we obtain f(n)= x(n)+Kx(n 1) b(n)= x(n 1)+Kx(n). Comparing with the prediction error: e(n)= x(n)ax(n 1) we identify K as -a. z -1 + + x(n) -a -a f(n)=e(n) g(n) Figure 13.17-1: 13.18 1 bkryy(lk)= rdy(l)= rxy(l),l =0,1 k=0 where y(n)is the input of the adaptive FIR filter B(z) ryyrssrwwrssw (lk)=(lk)+(lk)=(lk)+s2 d(lk) where s(n)is the output of the system C(z). If x(n)is white with variance s2 then, x rss(lk)=

s2 x (-0.9)|l-k|= s2 x (-0.9)|l-k| 10.9210.19 rxy(l)= E[x(n)y * (n l)]= E[x(n)(s * (n l)+w * (n l))]. If x(n)and w(n)are uncorrelated then, rxy(l)= E[x(n)s * (n l)]= sx2d(l). 406 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Thus, we obtain the system: . . 2 x + s2 w -2 x (0.9) s2 x b0 0.19 0.19 . . = b1 0 . 2 x 2 x

+ s2 w (0.9) 0.19 0.19

With s2 x and s2 w known, we can determine b0, b1. 13.19 (a) fm(n) = fm-1(n)kmgm-1(n 1) gm(n) = gm-1(n 1)k * mfm-1(n) eLS m = n . l=0 w n-l . |fm(l)|2 + |gm(l)|2. deLS m dk* m = -2 n . l=0 w n-l . g * m-1(l1)fm(l)+fm-1(l)g * m(l). = 0 n . n-l . * . * . wgm-1(l1)[fm-1(l)kmgm-1(l1)]+fm-1(l)gm-1(l1)kmfm-1(l) l=0 Solving for kM, we obtain 2Pn * n-lfm-1(l)gm-1(l1) um(n) l=0 w km(n)= 2. = Pn l=0 wn-l . |fm-1(l)|2 + |gm-1(l1)|vm(n) (b)

* wum(n 1)+2fm-1(n)gm-1(n 1) km(n)= 22 wvm(n 1)+|fm-1(n)|+ |gm-1(n 1)| fm-1(n)gm-1(n 1) = fm-1(n)g (n)+km(n)f * . m* m-1(n). * 2 = fm-1(n)g (n)+km(n)|fm-1(n)| m fm-1(n)gm-1(n 1) = gm-1(n)[fm(n)+km(n)gm-1(n 1)] 2 = gm-1(n 1)fm(n)+km(n)|gm-1(n 1)| Therefore, * 22 2fm-1(n)gm-1(n 1) = km(n) |fm-1(n)|+ |gm-1(n)|+ z(n) where * z(n)= fm-1(n)gm(n)+fm(n)gm-1(n 1) Now, * 22 2fm-1(n)gm-1(n 1) = z(n)+km(n) wvm(n)+|fm-1(n)|+ |gm-1(n)| -km(n)wvm(n 1) = z(n)+km(n)vm(n)km(n)wvm(n 1) 407 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N

o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Then, n-1 ** 2 w n-lfm-1(l)gm-1(l1)+2fm-1(n)gm-1(n 1) = wum(n 1) l=0 +z(n)+km(n)vm(n) -km(n)wvm(n 1) But km(n)= um(n)/vm(n). Therefore, km(n)wm(n)= z(n)+wum(n 1)+km(n)vm(n)kmwvm(n 1) and, then wum(n 1) z(n) km(n)= + wvm(n 1) wvm(n 1) z(n) km(n)= km(n 1)+ wvm(n 1) 408 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Chapter 14 14.1 (a) . T0 1 limT0.8E | x(t)e -j2Ftdt|2 2T0 -T0 . T0 . T0 -j2Ftdt j2Fdt = limT0.8E 1 x(t)ex * (t)e 2T0 -T0 -T0 . T0 . T0 = limT0.8 1 E[x(t)x * (t)]e -j2F(t-)dtdt 2T0 -T0 -T0 . T0 . T0 = limT0.8 1 .xx(t t)e -j2F(t-)dtdt 2T0 -T0 -T0

. t+T0 . T0 = limT0.8 1 .xx(a)e -j2F()dtda 2T0 t-T0 -T0 . 8 = .xx(a)e -j2F()da -8 = .xx(F) (b) N-1 .xx(m)= 1 . x(n + m)x * (n) N n=0 N NN-1 X. 1 X -j2fm -j2fm .xx(m)e = x(n + m)x * (n)e N m=-Nm=-Nn=0 N-1 n+N . 1 . -j2f(l-n) = x(l)x * (n)e N

n=0 l=n-N N-1N-1 1 X. -j2fl j2fn = x(l)x * (n)ee N n=0 l=0 N-1 1 . = | x(n)e -j2fn|2 N n=0 409 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14.2 N-|m|-1N-|m|-1 1 X. ' E[|.xx(m)|2]= N2 E[x * (n)x(n + m)x(n ' )x * (n + m)] n=0 n ' =0 1 X. ' = {E[x * (n)x(n + m)]E[x(n ' )x * (n + m)] N2 nn ' ' +E[x * (n)x(n ' )]E[x * (n + m)x(n + m)] ' +E[x * (n)x * (n + m)]E[x(n ' )x(n + m)]} 1 X. = N2 [.2 (m)+.2 (n n ' ) xxxx nn ' ' +. * (n +(n + m n ' )] xxm n).xx ' Let p = n -

n. Then 2 N -|m| 1 X. E[|.xx(m)|2]= .2 (m)+ [.2 (p). * (pm).xx(p+ m)] xxxxxx NN2 np 1 X. = |E[.xx(m)]|2 + xxxx(pm).xx N2 [.2 (p). * (p+ m)] np Therefore, 1 X. var[.xx(m)] = N2 [.2 (p). * (pm).xx(p+ m)] xxxx np 8

1 . [.2 (p). * (pm).xx(p+ m)] xxxx N p=-8 14.3 (a)

.. . N-|m|-1 . 1 . E[.xxxx. x * (n)x(n + m).. (m). * (m ' )] = E N . n=0 . .. N-|m|-1 1 X. ' x(n ' )x * (n + m ' ). . N . n ' =0 1 X. ' = E{x * (n)x(n + m)x(n ' )x * (n + m ' )} N2 nn ' 1 X. ' = {E[x * (n)x(n + m)]E[x(n ' )x * (n + m ' )]

N2 ' nn ' +E[x * (n)x(n ' )]E[x * (n + m ' )x(n + m)] ' +E[x * (n)x * (n + m ' )]E[x(n ' )x(n + m)]} x X. ' ) = Ns42 [d(m)d(m ' )+d(n n ' )d(m m ' nn '' +d(n + m n)d(n + m n ' )] N-1 N-1 -j2mf1-j2m ' f2 Hence, E[pxx(f1)pxx(f2)] = E[.xx(m).xx(m ' )]ee ' m=-(N-1) m =-(N-1) 410 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

x XXX. = Ns42 [d(m)d(m ' )+d(n n ' )d(m m ' ) mm ' nn ' '' -j2mf1-j2m ' f2 +d(n + m n)d(n + m n ' )]ee . 2 2) sinp(f1 + f2)N sinp(f1 f2)N = s4 1+ + x Nsinp(f1 + f2) Nsinp(f1 f2) (b) cov[pxx(f1)pxx(f2)] = E[pxx(f1)pxx(f2)]E[pxx(f1)]E[pxx(f2)] = E[pxx(f1)pxx(f2)]sx 4 2 2. s4 (sinp(f1 + f2)N sinp(f1 f2)N = x + Nsinp(f1 + f2) Nsinp(f1 f2) (c) var[pxx(f)] = cov[pxx(f1)pxx(f2)]|f1=f2=f "2# sin2pfN

s4 = 1+ x Nsin2pf 14.4 Assume that x(n) is the output of a linear system excited by white noise input w (n), where s2 = 1. Then pxx(f)=..xx(f)pww(f). From prob. 12.3,(a),(b) and(c), we have x E[pxx(f1)pxx(f2)] =..xx(f1)..xx(f2)E[pww(f1)pww(f2)] . 2 2) sinp(f1 + f2)N sinp(f1 f2)N =..xx(f1)..xx(f2)1+ + Nsinp(f1 + f2) Nsinp(f1 f2) cov[pxx(f1)pxx(f2)] =..xx(f1)..xx(f2)cov[pww(f1)pww(f2)] 2 2) (sinp(f1 + f2)N sinp(f1 f2)N =..xx(f1)..xx(f2) + Nsinp(f1 + f2) Nsinp(f1 f2) var[pxx(f)] = cov[pxx(f1)pxx(f2)]|f1=f2=f "2# sin2pfN =..f 1+ xx Nsin2pf 14.5 Let yk(n)= x(n)* hk(n)

N-1 j2k(n-m) N = x(m)e m=0 N-1 j2kn -j2km NN = ex(m)e m=0 N-1 -j2km N yk(n)|n=N = x(m)e m=0 = X(k) 411 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

NotethatthisisjusttheGoertzelalgorithmfor computingtheDFT.Then, N-1 -j2km N |yk(n)|2 = |X(k)|2 = | x(m)e |2 m=0 14.6 From(14.2.18) we have M-1 1 . W(f)= | w(n)e -j2fn|2 MU n=0 M-1M-1 1 X. -j2f(n-n ) = w(n)w * (n ' )e ' MU n=0 n ' =0 . 11 ' -j2f(n-n 2 W(f)df = 1 X. w(n)w * (n ' ) 2 e )df -1 MU -1

' 2 nn 2 1 X. = w(n)w * (n ' )d(n n ' ) MU ' nn . M-1 . 11 . = |w(n)|2 =1 UM n=0 by the definitionofUin(14.2.12) 14.7 (a)(1)Divide x(n)into subsequences of length M and overlapped by 50% to produce 4k subse 2 quences. Each subsequence is padded with M zeros. 2 (2)Compute the M-point DFT of each frame or subsequence. (3)Compute the magnitude square of each DFT. (4)Average the 4k M-point DFT s. (5)Perform the IDFT to obtain an estimate of the autocorrelation sequence. (b) M-1 j2km M

X3(k)= x3(m)e m=0 M -1 2 M-1 . -j2km . M j2km M M = x1(m)e + x2(m )e 2 M m=0 m= 2 M-1 M-1 . j2km . j2km ' --jk = x1(m)e M + ex2(m ' )e M m=0 m ' =0 -jkX2(k) X3(k)= X1(k)+e (c) Instead of zero-padding, we can combine two subsequences to produce a single M-point subsequence and thus reduce the number of sequences form 4k to 2k. Then, we use the relation in(b) for the DFT. 412 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14.8 0.9 (a)Obviously, .f =0.01. From(12.2.52), M = =90. .f (b) From(14.2.53), thequalityfactor is QB =1.1N.f. This expression does not depe nd on M; hence, there is no advantage to increasing the value of M beyond 90. 14.9 (a)From table 14.1, we have QB =1.11N.f QB 1 ..f == 1.11N 111 Qw =1.39N.f Qw 1 ..f == 1.39N 139 QBT =2.34N.f QBT 1 ..f == 2.34N 234 (b) For the Bartlett estimate, N QB = M N . M = =100 QB For the Welch estimate with 50% overlap, 16N Qw = M

16N . M = =178 Qw For the Blackman-Tukey estimate, 1.5N QBT = M 1.5N . M = =150 QBT 14.10 (i) (a)SupposeP(f)is the periodogram based on the Bartlett method. Then, B P(i) B (f) = 1 M| M-1 . n=0 xi(m)e -j2fn|2 , i = 0,1,. . . ,k 1 P(0) xx (f) = 0 P(1) xx (f) = 1w M | M-1 . n=0 x1(m)e -j2fn|2 = (1w)P(1) B (f) P(2) xx (f) = wP(1) xx (f)+ (1w)P(1) B (f) 413 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim

itris G. Manolakis. ISBN 0-13-187374-1.

(1) (2) = (1w)[wP(f)+P(f)] BB P(m) . m-kP(k) (f) =(1w) mw (f) xx B k=1 . (k) Therefore, E{P(M)(f)} = (1w) Mwm-kE[P(f)] xx B k=1 M Z 1w1 1 2 = 2 sinp(f a)M (1w) ..xx(a) da 1wM sinp(f a) 1 2 1 2 2

sinp(f a)M (1w M) 1 ..xx(a) da = M sinp(f a) 1 2 var{P(M)(f)} = E{[P(M)(f)]2}[E{P(M)(f)}]2 xx xxxx . (k) var{P(M)(f)} = E{[(1w) Mwm-kP(f)]2} xx B k=1 . (k) -{E[(1w) Mwm-kP(f)]}2 B k=1 . (k)(k) = (1w)2 Mw2(M-k)E{P(f)}2 -{E[P(f)]}2 BB k=1 (k) = (1w)2 . Mw2(M-k)var[P(f)] B k=1

2M 2. 1w"sin2pfM = (1w)2..2 (f) 1+ xx 1w2 Msin2pf 2. 1w "sin2pfM 2w) = (1w ..2 (f) 1+ xx 1+w Msin2pf (b) E{P(M) xx (f)} = 1 2 E{P(w) xx (f)} =..xx(a)W(f a)da 1 2 M-1 1 .

where W(f)= | w(n)e -j2fn|2 MU n=0 M . P(i) var[P(M)(f)] =(1w)2 w2(M-k)var[ (f)] xx xx k=1 1w . 2M)..2 = (1w (f) xx 1+w 14.11 Let R(i) xx be defined as follows: . r(i) xx (0) r(i) xx (1) . . . . R(i) xx = 1 M . . . . . r(i) xx (-1) r(i) xx (0) . . . . . . . . . . . r(i) xx (0) 414

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Then, M-1M-1 E *t(f)R(i) X. 1(kk ' )e (i) -j2(k-k ' )f E(f)= r xxxx M k=0 k' =0 M-1 1 . = M k=0 (M-1) = -(M-1) P(i) =(f) xx K 1 . (i) -j2mf r(m)e xx m=k-(M-1) (M|m|) (i) -j2mf r(m)e xx

M Therefore, P(B)(f)= E *t(f)R(k)E(f) xx xx K k=1 14.12 Toprovetherecursiverelationin(12.3.19) wemake useofthefollowingrelations: E m = N-1 . n=m [|fm(n)|2 + |gm(n 1)|2] (1) where fm(n) = fm-1(n)+kmgm-1(n 1) gm(n) = k * mfm-1(n)+gm-1(n 1) (2) and E m-1 = N-1 . n=m-1 [|fm-1(n)|2 + |gm-1(n 1)|2] = |fm-1(m 1)|2 + |gm-1(m 2)|2 + N-1 . n=m [|fm-1(n)|2 + |gm-1(n 1)|2] Also, N-1 . n=m [fm-1(n)+g * m-1(n 1)] = 1 2 km E m-1 We substitute for fm(n)and gm(n 1) from(2) into(1),andwe expandthe expressions. Then, use the relations for E m-1 and k m to reduce the result.

14.13 1 x(n)= x(n 1)+w(n)w(n 1) 2 1 E[x(n)] = E[x(n 1)]+E[w(n)]E[w(n 1)] 2 since E[w(n)] =0, it follows that E[x(n)]=0 To determine the autocorrelation, we have 1 h(0) = h(-1)+d(0)d(-1) = -1 2 11 h(1) = h(0)+d(1)d(0)= 22 415 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

1 p= q =1,a = ,b0 =1,b1 = -1 2Hence, .xx(0) = .xx(1) = and .xx(0) = .xx(1) = .xx(m)= = .xx(m)= = 14.14 x(n) E[x(n)] .xx(m) where q =2,b0 =1,b1 = -2,b2 =1 Hence, .xx(0) .xx(1) .xx(2) .xx(m) .xx(-m) 14.15 (a) ..xx(z)= = vv 1 j 31 j 3 The four zeros are , 22 The minimum-phase system is H(z)= Hence, H(z)= 416 1 .xx(1)+s2 (1+ 1

) w 22 1 .xx(0)+s2 (-1) w 2 4 s2 w 3 1 s2 w 3 -a1.xx(m 1) 1 1)m-1s2 ( , m> 1 w 32 .xx(-m) 1 1)-m+1s2 ( , m< 0 w

32 = w(n)2w(n 1)+w(n 2) = 0 since E[w(n)]=0 q = s2 0 = m = q w bkbk+m, k=0 2 = s2 . b2 = bs2 wk w k=0 2 = s2 . bkbk+1 = -4s2 ww k=0 2 s2 . s2 = w bkbk+2 = w k=0 =0, |m|= 3, = .xx(m) -m .xx(m)z m 2z -2(z 4 2z 3 +3z 2 2z +1) v -1

G(1z + z -2), where G =2 v -1 -2) 2(1z + z (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) The solution is unique. 14.16 (a) 8 ..xx(z) = . .xx(m)z -m m=-8 = z2 62(635z -1 + 62z -2 35z -3 + 6z -4) = z2 62(13z -1)(12z -1)(11 2 z -1)(11 3 z -1) 1 1 The four zeros are z = 3,2, 3, 2 The minimum phase system is H(z) = 6 v 62(11 2 z -1)(11 3 z -1) = 1 v 62(65z -1 + z -2) v1 -2) (b) The maximum phase system isH(z)= (15z-1 +6z 62v1 -2) (c) There are two possible mixed-phase systems: H1(z)= (37z-1 +2zH2(z)= 62v1 (27z-1 +3z-2)

62 14.17 (a) -1 1+z H(z)= 10.8z-1 ..hh(f)= H(z)H(z z=e -1)| j2f -j2f j2f 1+e1+e = 10.8e-j2f 10.8ej2f cos2pf =4 1.641.6cos2pf 1)|m| .xx(m) =( 2 8 . 1)|m|-j2fm . ..xx(f) =( e 2 m=-8 0.75 = 1.25cos2pf ..yy(f) =..xx(f)..hh(f)

3cos2pf = (1.641.6cos2pf)(1.25cos2pf) (b) 75 54 2 ..yy(f)= 1.641.6cos2pf 1.25cos2pf 417 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

93 25 4 = 150 50 1.641.6cos2pf 1.25cos2pf 1)|m| .yy(m) = 150(0.8)|m|50( 2 (c)s2 = (0)=15050 =100 w .xx 14.18 proof is by contradiction. (a) Assume the |km| > 1. Since Em = (1-|km|2)Em-1, this implies that either Em < 0 or Em-1 < 0. Hence, s2 < 0, and w . s2 w 0 . . . 0 . .. ...xx . ... . t

t.. aa = a xx is not positive definite. (b) From the Schur-Cohn test,Ap(z)is stable if |km| < 1. Hence, the roots of Ap(z)are inside the unit circle. 14.19 (a) . . . .. . s2 w .xx(0) .xx(1) .xx(2) 1 .xx(-1) .xx(0) .xx(1) 0 0 . . . . = . . .xx(-2) .xx(-1) .xx(0) -0.81 0 .xx(m) =0.81.xx(m 2),m = 3

.xx(m) Hence, = {2.91,0,2.36,0,1.91,0,1.55,0,...} s2 w q The values of the parameters dm = bkbk+m are as follows: k=0 MA(2): dm = {2.91,0,2,36} MA(4): dm = {2.91,0,2,36,0,1.91} MA(8): dm = {2.91,0,2,36,0,1.91,0,1.55,0} (b) The MA(2),MA(4)and MA(8) models have spectra that contain negative values. O n the otherhand,thespectrum oftheARprocessisshownbelow. Clearly,theMAmodelsdo not provide good approximations to the AR process. Refer to fig 14.19-1. 14.20 .xx(m) = 1.656s2 ,0,0.81s2 ww ,0, ... . ForAR(2) process: 418 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 1 2 3 4 5 6 ---> magnitude 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 1 2 3 4 5 6 ---> magnitude ---> frequency(Hz) Figure 14.19-1: 1.656s2 w 0 . 0 1.656s2 w 0.81s2 w 0 The solution is g a1 a2 . FortheAR(4) process,weobtain g a FortheAR(8) process,weobtain g a Refer to fig 14.20-1. 14.21 (a)(1) H(w)= ..xx(w)= ..xx(w)= (2) H(w)

..xx(w) (3) H(w)= 0.81s2 w . . 1 . . gs2 w . 0 . . a1 . = . 0 . 1.656s2 w a2 0 = 1.12 = 0 = -0.489 = 1.07 and = {1,0,-0.643,0,0.314} =1.024 and = {1,0,-0.75,0,0.536,0,-0.345,0,0.169} -jw 1e1+0.81e-jw |H(w)|2s2 w -jw 1e ||2s2 w 1+0.81e-jw -j2w) = (1e = |H(w)|2s2 w =4s2 sin2 w w 1

10.81e-jw 419 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

MA(2) AR(2) 2 2.5 1.5 2 1 1.5 ---> magnitude ---> magnitude ---> magnitude ---> magnitude 0.5 1 0 0.5 0 0.2 0.4 0.6 0 0.2 0.4 0.6 ---> frequency(Hz) ---> frequency(Hz) AR(4) AR(8) 2 2 1.5 1 0.5 1.5 1 0.5 0 0 0.2 0.4 0.6 0 0.2 0.4 0.6 ---> frequency(Hz) ---> frequency(Hz) Figure 14.20-1:

s2 w ..xx(w)= 1.65611.62cosw (b) Refer to fig 14.21-1. (c) For(2), .P3 s2 0 = m = 2 . wk=0 bkbk+m, .xx(m)= 0, m> 2 . .* (-m), m< 0 xx since b0 =1,b1 =0 and b2 = -1, we have .xx(0) = w 2s2 .xx(2) = w -s2 (-2) = -s2 .xx(m) =0,m 6=0,2 .xxw For(3),theAR process has coefficients a0 =1,a1 =0 and a2 =0.81. . .. ... s2 1 00.81 .xx(0) w . 01.81 0 .. .xx(1) . = . 0 . 0.81 0 1 .xx(2) 0 (0) =2.9s2 .xxw

420 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(1) (2) 8 2 6 4 ---> magnitude ---> magnitude 0.5 1 1.5 0 0.2 0.4 0.6 ---> magnitude 2 0 0 0.2 0.4 0.6 ---> frequency(Hz) (3) 6 5 4 3 2 1 0 0 0.2 0.4 0.6 ---> frequency(Hz) 0 ---> frequency(Hz) Figure 14.21-1: .xx(m) =0,m odd .xxw (m) =2.9(0.9)|m|s2 ,m even 14.22 (a)For the Bartlett estimate, 0.9 M = .f 0.9 = = 90 0.01 0.9

(b)M = 0.02 = 45 N (c)for (a),QB = M 2400 = = 26.67 90 N for (b), QB = M 2400 = = 53.33 45 421 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14.23 |ej2f 0.9|2 ..xx(f)= s2 w |ej2f j0.9|2|ej2f + j0.9|2 (a) z 0.9 z-1 0.9 (z)= s2 ..xxw z2 +0.81 z-2 +0.81 z 0.9 Therefore, H(z)= z2 +0.81 z-1(10.9z-1) = 1+0.81z-2 (b) The inverse system is -2 1 1+0.81z = H(z) z-1(10.9z-1) This is a stable system. 14.24 N-1 -j2nk X(k)= x(n)e N n=0 (a) -j2nk

E[X(k)] = E[x(n)]e N =0 n -j2k(n-m) N E[|X(k)|2]= E[x(n)x * (m)]e nm -j2k(n-m) N = s2d(n m)e x nm N-1 = s2 . 1 x n=0 = Ns2 x (b) E{X(k)X * (km)} = . n . n ' E[x(n)x * (n ' )]e -j2kn N e j2n ' (k-n) N = s2 x . . d(n n ' )e -j2mn ' N e -j2k(n-n ' ) N ' n n = s2 xe j2mn N = Ns2 x, m = pN =0, otherwise p=0,1,2,... 422

(text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14.25 .vv(m)= E[v * (n)v(n + m)] qq = k' =0 k=0 b * kbk' E[w * (n k)w(n + m k ' )] qq s2 w b * kbk' d(m + kk ' ) = k' =0 k=0 q s2 w b * kbk+m = k=0 = s2 dm w q = s2 w dme -j2fm Then, ..vv(f) m=-q

14.26 .xx(m)= E[x * (n)x(n + m)] = A2E{cos(w1n + f)cos[w1(n + m)+f]} A2 = E{cosw1m + cos[w1(2n + m)+2f]} 2 A2 = cosw1n 2 14.27 (a) x(n) =0.81x(n 2)+w(n) y(n)= x(n)+v(n) . x(n)= y(n)v(n) y(n)v(n) =0.81y(n 2)0.81v(n 2)+w(n) Therefore, y(n) =0.81y(n 2)+v(n)0.81v(n 2)+w(n) so that y(n)is an ARMA(2,2) process (b) p x(n)= akx(n k)+w(n) k=1 y(n)= x(n)+v(n) . x(n)= y(n)v(n) p

y(n)v(n)= ak[y(n k)v(n k)]+w(n) k=1 pp y(n)+ aky(n k)= v(n)+ akv(n k)+w(n) k=1 k=1 423 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Hence, y(n)is an ARMA(p,p) process p -k] Note that X(z)[1+ akz = W(z) k=1 H(z) ..xx(z) and ..yy(z) = = = = = = 1 1+Pp k=1 akz-k 1 Ap(z) s2 wH(z)H(z -1) s2 wH(z)H(z -1)+s2 v s2 w Ap(z)Ap(z-1) + s2 v s2 w + s2 vAp(z)Ap(z-1) Ap(z)Ap(z-1) 14.28 (a) K K .xx(m)= E{[ Akcos(wkn + fk)+w(n)][ Ak' cos(wk' (n + m)+fk' )+w(n + m)]} k=1 k' =1 = AkAk' E{cos(wkn + fk)cos(wk' (n + m)+fk' )} + E[w(n)w(n + m)] kk' K A2

cos(wkn)+s2 d(m) w = 2 k=1 (b) 8 -jwm ..xx(w)= .xx(m)e m=-8 K 8 A2 jwk + -jwk)e -jwn (ee + s2 w = 4 k=1 m=-8 K A2 [2pd(w wk 2pm)+2pd(w + wk 2pm)]+s2 w = 4 k=1 K p A2 k[d(w wk 2pm)+2pd(w + wk 2pm)]+s2 w

= 2 k=1 14.29 E = a *T ..yya + .(1a *T a) dE = 0 da . ..yya .a = 0 or ..yya = .a 424 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

Thus, a is an eigenvector corresponding to the eigenvalue .. Substitute ..yya = .a into E. Then, E = .. To minimize E, we select th smallest eigenvalue, namely, s2 . w 14.30 (a) .xx(0) .xx(1) .xx(2) By the Levinson-Durbin algorithm, = = = a1(1) = = k1 E1 = = = a2(2) = = a2(1) = = (b)k2 = a2(2) k1 = a1(1)as given above. (c) If s2 . w a2(1) = = a2(2) = k2 = k1 = 14.31 P + s2 w Pcos2pf1 Pcos4pf .xx(1) -

.xx(0) Pcos2pf1 P + s2 w a1(1) (1k12).xx(0) P2sin22pf1 +2Ps2 + s4 ww P + s2 w .xx(2)+a1(1).xx(1) E1 Ps2 cos4pf1 P2sin22pf1 w P2sin22pf1 +2Ps2 + s4 ww a1(1)+a2(2)a1(1) Pcos2pf1 . P2sin22pf1 Ps2 cos4pf1 . w 1+ P + s2 P2sin22pf1 +2Ps2 + s4 w ww 0, we have

-(cos2pf1)(1+1) -2cos2pf1 1 1 -cos2pf1 e(h)= hH..xxh+ (1EH(f)h)+ * (1hHE(f)) (a)To determine the optimum filter that minimizessy 2 subject to the constraint, we differentiate e(h)with respect to hH (compute the complex gradient): e(h) =..h* E(f)=0 xx hH Thus, * ..-1 h= E(f) opt xx 425 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

. (b) To solve for the Langrange multipliers using the constraint, we have * EH EH(f)h= (f)..-1E(f)=1 opt xx Thus, 1 * = EH(f)..-1E(f) xx * By substituting for ..-1 E(f) xx = hopt EH(f)..-1E(f) xx 14.32 The periodogram spectral estimate is 11 X(f)X * (f) 2 PXX(f)= |X(f)|= NN where N-1 . intheresultgivenin(a) weobtaintheoptimum filteras

X(f)= x(n)e -j2fn EH = (f)X(n) n=0 By substituting X(f)into Pxx(f), we obtain 1(f)X(n)X(n)HE(f) EH Pxx(f)= N Then, 1 EH (f)E X(n)X(n)H E[Pxx(f)] E(f) = N 1 EH (f)..E(f) xx = N 14.33 WeusethePisasenko decompositionmethod.First,wecomputetheeigqnvaluesofthecorrelat ion matrix. 3. 0 -2 03. 0

-2 03. 2 03. 3. 0 3. 0 03. g(.) (3.) = = (3.)3 2(2)(3.) (3.)2 4 (3.) 0 = = = Thus, . =5,3,1 and the noise varinace is .min =1. The corresponding eigenvector is . .. .. . .

. 20 -2 1 0 1 020 0 . . a2 =1,a1 =0 . 0 a1 . = . . .. . . -20 2 a2 01 -2 jw The frequency is found from the equation 1+ z=0 . z = j. Therefore, e= j yields w = p/2 and the power is P =2. 426 (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved . This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

14.34 The eigenvalues are found from g(.) = 2. -j -1 j 2. -j -1 j 2. . .1 =1,.2 =1,.3 =4. and the normalized eigenvectors are v -j/ 3 v . . . . . . 0 v 2/3 v j/ 2 v 1/ 3 v

-j/ 6 v v1 = . . v2 = . . v3 = . . j/ 31/ 21/ 6 Bycomputingthe denominatorof(14.5.28),we findthatthefrequencyis . = p/2or f =1/4 . We may also find the frequency by using the eigenvectors v2 and v3 to constructthet wopolynomials (Boot Music Method): j1 -2 V2(z)= v z -v z 22 21 1 -1 -2 -v z + v z 36 6 V3(z) = Then, we form the polynomials 12 21

** 2 -2 V2(z)V2 (1/z * )+V3(z)V3 (1/z * )= z + jz+2 jz-1 + z 3333 Itis easilyverifiedthatthepolynomialhasadoubleroot at z = jor, equivalently, at . = p/2. The other two roots are spurious roots that are neglected. Finally, the power of the exponential signal is P1 =1. 14.35 1 PMUSIC(f)= PM 2 |sH(f)vk| k=p+1 The denominator can be expressed as MM s H(f)vk 2 = s H H(f)vkvk s(f) k=p+1 k=p+1 . . M

. vkv H . s(f) k = s H(f) k=p+1 14.36 PM-1 (a)Vk(z)= vk(n+1)z-n and Vk(f)= Vk(z)|j2f Then,the denominatorinPMUSIC(f) n=0 z=e may be expressed as M = s H(f)vk 2 * MVk(f)Vk (f) = k=p+1 k=p+1 = k=p+1 427 * MVk(z)Vk (1/z * )| z=ej2f (text) 2007 Pearson Education, Inc., Upper Saddle River, NJ. All rights reserved

. This material is protected under all copyright laws as they currently exist. N o portion of this material may be reproduced, in any form or by any means, witho ut permission in writing from the publisher. For the exclusive use of adopters o f the book Digital Signal Processing, Fourth Edition, by John G. Proakis and Dim itris G. Manolakis. ISBN 0-13-187374-1.

(b) For the roots ofQ(z),we consruct(fromProblem14.34)Q(z)as ** Q(z)= V2(z)V2 (1/z * )+V3(z)V3 (1/z * ) 12 21 2 -2 = z + jz+2 jz-1 + z 3333 Thus polynomial has a double root at z = j and two spurious roots. Therefore, th e desired frequency is . = p/2. 14.37 (a) N-1 .xy(n0)= y(n n0)[y(n n0)+w(n)] n=0 N-1 E[.xy(n0)] = E[y 2(n n0)] n=0 N-1 2 = E[A2 cos w0(n n0)] 0 = n = M 1 n=1 MA2 = 2

MA2 var[.xy(n0)] var[.xy(n0)] = E[.2 xy(n0)]( 2 )2 = . . n n ' E{y(n n0)[y(n n0)+w(n)]y(n ' n0)[y(n ' n0)+w(n ' )]} ( )2MA2 = MA2 2 s2 w (b) SNR = {E[.xy(n0)]}2 (MA2 )2 2 = MA2 s2 2 w MA2 = 2s2 w (c) As M increases, the SNR increases. 14.38 Refer to fig 14.38-1.