TUGAS PENGOLAHAN SINYAL DIGITAL

KELOMPOK 1

Oleh :

Nikko Haendratnio

H1C007061

Henky Agung W.

H1C008066

Muhamad Abduh

H1C009003

Suwitno

H1C009039

KEMENTERIAN PENDIDIKAN NASIONAL

UNIVERSITAS JENDRAL SOEDIRMAN
FAKULTAS SAINS DAN TEKNIK
JURUSAN TEKNIK
PROGRAM STUDI TEKNIK ELEKTRO
PURBALINGGA
2011/2012

TUGAS PSD CHAPTER 4

1. given a sequence x(n) for 3 where x(0) = 1, x(1) = 1, x(2) = -1 and x(3) = 0, compute its DFT
X(k).
Answer:
X(0) = = ( x(0) + x(1) + x(2) + x(3) ) = (1+1-1+0) = 1
X(1) = = ( + + )
= x(0) - jx(1) x(2) + jx(3) = 1 j + 1 + 0 = 2-j
X(2) = = ( + + )
= x(0) - x(1) + x(2) x(3)
=111+0 =-1
X(3)

= =( + +)

= x(0) + jx(1) - x(2) jx(3)

=1+j+1-0=2+j

2. Given a sequence x(n) for 0 <= 3, where x(0) = 4, x(1) = 3, x(2) = 2,

and x(3) = 1, evaluate its DFT X(k).
N

=4

W4

x(n) W4 kn

X(k) =

n=0
3
=

x(n)

kn
2

n=0
untuk :
3
k = 0 = x(0) = x(n)
x(3)

e j0 = x(0) e j0 + x(1) e j0 + x(2)

j0

n=0
= x(0) + x(1) + x(2) + x(3)
=4+3+2+1

e j0 +

= 10
3
j

k = 1 = x(1) = x(n)
x(3)

n
2

= x(0)

j0

+ x(1) + e j 2 + x(2)

3
2

n=0
= x(0) - jx(1) + x(2) + jx(3)
= 4 j3 2 + j
= 2- j2

3
k = 2 = x(2) = x(n)
j3

e j n = x(0)

e j0 + x(1)

e j + x(2) e j 2 + x(3)

n=0
= x(0) - x(1) + x(2) - x(3)
=43+2-1
=2
3
k = 3 = x(3) = x(n)

3 n
2

= x(0)

j0

+ x(1)

j3

n=0
+ x(3)

9
2

= x(0) + j x(1) - x(2) - jx(3)

= 4 + j3 - 2 j
= 2 +j2
dengan menggnakan MATLAB function fft(), maka :
X = fft([1 1 -1 0])
X = 1.0000
2.0000 - 1.0000i -1.0000
2.0000 + 1.0000i

3
2

+ x(2)

3. Given the DFT sequence X(k) for 0k 3 obtained in Problem 4.2, evaluate its inverse
DFT x(n).
jawab:

N = 4 dan W4-1 = e j 2 , maka:

x n=

k n
j
1
X k W4-nk = x n= 1 X k e 2

4
4

untuk n = 0
x 0=
=

1
1
X 0e j 0= X 0e j 0 X 1e j 0 X 2e j 0 X 3 e j 0
4
4
1
10 2J2 2 2J2 =4
4

untuk n = 1
j
1
x 1= X k e
4

k
2

j
j
1
= X 0 e j 0 X 1 e 2 X 2e j X 3e
4

1
X 0 jX 1 X 2 jX 3
4

1
10 j 2 j22 j 2 j2=3
4

3
2

untuk n = 2
x 2=

1
1
X k e j k = X 0e j 0 X 1e j X 2 e j 2 X 3 e j 3
4
4

1
X 0 X 1 X 2 X 3
4

1
10 2 j222 j2=2
4

untuk n = 3
j
1
x 3= X k e
4

k 3
2

j
1
= X 0 e j 0 X 1 e
4

1
X 0 jX 1 X 2 jX 3
4

1
10 j 2 j22 j 2 j2=1
4

3
2

X 2 e

j 3

X 3 e

33
2

4.

Penyelesaian :

Since N = 4 and W4 = e j 2

We have simplified formula :

n=0

x nW4

X(k) =

n=0

kn

x n

This , for k=0

n=0

X(0)

x n

j0

j0
j0
j0
j0
= x(0) e
+ x(1) e
+ x(2) e
+ x(3) e

= x(0) + x(1) + x(2) + x(3)

=4+3+2+1
= 10
for k=1
n=0

X(1)

x n

n
2

j0

= x(0) e

+ x(1) e

+x(2) e

+x(3) e

3
2

= x(0) - jx(1) - x(2) + jx(3)

= 4 - j3 - 2 + j
= 2 - 2j
for k=2
n=0

X(2)

x n
3

e j n

j0
j
j2
j3
= x(0) e
+ x(1) e
+x(2) e
+x(3) e

= x(0) - x(1) + x(2) - x(3)

=4-3+21
=2
and for k=1

n=0

X(3)

x n

3n
2

j0

= x(0) e

+ x(1) e

3
2

j3

+x(2) e

+x(3) e

9
2

= x(0) + jx(1) - x(2) - jx(3)

= 4 + j3 - 2 - j
= 2 + 2j
DFT Using Matlab :
>> X = fft ([4 3 2 1])
X=
10.0000

2.0000 - 2.0000i 2.0000

2.0000 + 2.0000i

DFT Using matlab with zero padding :

>> X = fft ([4 3 2 1 0 0])
X=
10.0000

3.5000-4.3301i 2.5000-0.8660i

2.0000

2.5000+0.8660i 3.5000+4.3301i

5. Using the DFT sequence X(k) for computed in problem 4.4, evaluate the inverse DFT x(0)
and x(4)
Answer:
Dari soal 4 didapatkan:

x(0) = 10
X(1) = 3,5 4,3301j
X(2) = 2,5 0,8660j
X(3) = 2
X(4) = 2,5 + 0,8660j
X(5) = 3,5 + 4,3301j

Ditanyakan x(0) dan x(4)

X(0)= (x(0) + x(1) + x(2) + x(3) + x(4) + x(5)
= ( 10 + 3,5 - 4,3301j + 2,5 0,8660j + 2 + 2,5 + 0,8660j + 3,5 + 4,3301j)

= (10 + 3,5 + 2,5 + 2 + 2,5 + 3,5)

= (24)
=4
X(4)=
= (+ + + + )
= ( 10 (3,5 - 4,3301j) (2,5 0,8660j) + 2 (2,5 + 0,8660j) (3,5 + 4,3301j))
= (0)
=0

6. Consider a digital sequence sampled at the rate of 20,000 Hz. If we use

the 8,000-point DFT to compute the spectrum, determine :
a. the frequency resolution
b. the folding frequency in the spectrum.
Answer:
a.

f =

b. f max =

fs
=
N

N
2

20.000
= 2.5 Hz
8000

f = 10.000 Hz

7. We use the DFT to compute the amplitude spectrum of a sampled data sequence with a
sampling rate fs = 2,000 Hz. It requires the frequency resolution to be less than 0.5 Hz.
Determine the number of data points used by theFFTalgorithm and actual frequency
resolution inHz, assuming that the data samples are available for selecting the number of
data points.
Jawab: f =0.5 Hz
N=

fs 2000
=
=4000
f
0.5

Since we use the FFT to compute the spectrum, the number of the data points must be a

power

of 2, that is,

N =215=32768
And the resulting frequency resolution can be recalculated as
f=

fs 2000
=
=0.061 Hz
N 32768

8. Penyelesaian :
Diketahui

N=4

fs = 100 Hz

W4 = e j 2

x(0) = -1; x(1)= 2; x(2) = 1; x(3) = 4

Ditanyakan

: amplitude spectrum, phase spectrum and power spectrum . . .?

Jawab

using matlab function we can get DFT X(k)

>> X=fft ([-1 2 1 4])
X=
6.0000

-2.0000 + 2.0000i -6.0000

-2.0000 2.0000i

we get X(0) = 6;
X(1)

= -2+2j = 2.828 <1350;

X(2)

= -6;

X(3)

= -2-2j = 2.828 <-1350

the amplitude spectrum, phase spectrum and power spectrum are computed as follows
:For k = 0, f = k . fs /N = 0 x 100/4 = 0 Hz
A0 =

1
|X(0)| =
4

0 = tan-1 ( |
P0 =

1
x 6 = 1.5 ;
4
ImagX 0
0
| ) = tan-1 ( |
| ) = 00
6
RealX 0

1
|X(0)|2 =
42

1
|6|2 = 2.25
16

For k = 1, f = k . fs /N = 1 x 100/4 = 25 Hz
A1 =

1
|X(1)| =
4

1
x 2.828 = 0.707;
4

1 = tan-1 ( |

P1 =

ImagX 0
2
| ) = tan-1 ( |
| ) = 1350
2
RealX 0

1
2
2 |X(1)| =
4

1
|2.828|2 = 0.5
16

For k = 2, f = k . fs /N = 2 x 100/4 = 50 Hz
A2 =

1
|X(2)| =
4

2 = tan-1 ( |

P2 =

1
x (-6) = -1.5 ;
4

ImagX 0
0
| ) = tan-1 ( |
| ) = 00
6
RealX 0

1
|X(2)|2 =
42

1
|-6|2 = 2.25
16

For k = 3, f = k . fs /N = 3 x 100/4 = 75 Hz
A3 =

1
|X(3)| =
4

3 = tan-1 ( |

P3 =

1
x 2.828 = 0.707 ;
4

ImagX 0
2
| ) = tan-1 ( |
| ) = -1350
2
RealX 0

1
|X(3)|2 =
42

1
|2.828|2 = 0.5
16

9. Compute the following window functions for size of 8:

Hamming window function
Hanning window function
Answer :
N=8
Whm(0) = 0,54 - 0,46 cos () = 0.0800
Whm(1) = 0,54 0,46 cos () = 0,2532
Whm(2) = 0,54 0,46 cos () = 0,6424
Whm(3) = 0,54 0,46 cos () = 0,9544
Whm(4) = 0,54 0,46 cos () = 0,9544
Whm(5) = 0,54 0,46 cos () = 0,6424
Whm(6) = 0,54 0,46 cos () = 0,2532
Whm(7) = 0,54 0,46 cos () = 0,0800

Jadi fungsi jendela hamming adalah :

W = [ 0,0800 0,2532 0,6424 0,9544 0,9544 0,6424 0,2532 0,0800 ]
N=8 dicari fungsi jendela hanning
Whn(0) = 0,5 0,5cos () = 0
Whn(1) = 0,5 0,5cos () = 0,1883
Whn(2) = 0,5 0,5cos () = 0,6113
Whn(3) = 0,5 0,5cos () = 0,9505
Whn(4) = 0,5 0,5cos () = 0,9505
Whn(5) = 0,5 0,5cos () = 0,6113
Whn(6) = 0,5 0,5cos () = 0,1883
Whn(7) = 0,5 0,5cos () = 0
Jadi fungsi jendela hanning adalah :
W = [ 0 0,1883 0,6113 0,9505 0,9505 0,6113 0,1883 0 ]

10. Given the following data sequence with a length of 6,

x(0) = 0, x(1) = 1, x(2) = 0, x(3) = 1, x(4) = 0, x(5) = 1
compute the windowed sequence xw(n) using the
a. triangular window function.
b. Hamming window function.
c. Hanning window function.
Answer:
N=6
a. triangular
wtri (n) = 1wtri (0) = 1-= 0
wtri (1) = 1-= 0.4

wtri (2) = 1-= 0.8

wtri (3) = 1- = 0.8
wtri (4) = 1- = 0.4
wtri (5) = 1- = 0

windowed sequence :
Xw (0) = X (0) x Wtri (0) = 0 x 0

=0

Xw (1) = X (1) x Wtri (1) = 1 x 0.4 = 0.4

Xw (2) = X (2) x Wtri (2) = 0 x 0.8 = 0
Xw (3) = X (3) x Wtri (3) = -1 x 0.8 = -0.8
Xw (4) = X (4) x Wtri (4) = 0 x 0.4 = 0
Xw (5) = X (5) x Wtri (5) = 1 x 0

=0

b. Hamming
Whm (n) = 0.54 0.46 cos
Whm (0) = 0.54 0.46 cos = 0.08
Whm (1) = 0.54 0.46 cos = 0.4
Whm (2) = 0.54 0.46 cos = 0.91
Whm (3) = 0.54 0.46 cos = 0.91
Whm (4) = 0.54 0.46 cos = 0.4
Whm (5) = 0.54 0.46 cos = 0.08

windowed sequence :
Xw (0) = X (0) x Whm(0) = 0 x 0.08

=0

Xw (1) = X (1) x Whm (1) = 1 x 0.4

= 0.4

Xw (2) = X (2) x Whm (2) = 0 x 0.91 = 0

Xw (3) = X (3) x Whm (3) = -1 x 0.91 = -0.91

Xw (4) = X (4) x Whm (4) = 0 x 0.4

=0

Xw (5) = X (5) x Whm (5) = 1 x 0.08

= 0.08

c. Hanning
whn (n) = 0.5- 0.5 cos
whn (0) = 0.5- 0.5 cos = 0
whn (1) = 0.5- 0.5 cos = 0.35
whn (2) = 0.5- 0.5 cos = 0.9
whn (3) = 0.5- 0.5 cos = 0.9
whn (4) = 0.5- 0.5 cos = 0.35
whn (5) = 0.5- 0.5 cos = 0
windowed sequence :
Xw (0) = X (0) x Whn(0) = 0 x 0

=0

Xw (1) = X (1) x Whn (1) = 1 x 0.35 = 0.35

Xw (2) = X (2) x Whn (2) = 0 x 0.9 = 0
Xw (3) = X (3) x Whn (3) = -1 x 0.9 = -0.91
Xw (4) = X (4) x Whn (4) = 0 x 0.35 = 0
Xw (5) = X (5) x Whn (5) = 1 x 0

=0

11.

a. Triangular Window
N = 4, maka:
Wtri(0) = 1

2 x 041
=0
41

Wtri(1) = 1

2 x 141
= 0.6667
41

Similarly, wtri(2) = 0:6667, wtri(3) = 0. Next, the windowed sequence is computed as

xw(0) = x(0) x Wtri(0) = 4 x 0 = 0
xw(1) = x(1) x Wtri(1) = -1 x 0.6667 = -0.6667
xw(2) = x(2) x Wtri(2) = 2 x 0.6667 = 1.3334
xw(3) = x(3) x Wtri(3) = 1 x 0 = 0
Applying DFT Equation (4.8) to xw(n) for k = 0, 1, 2, 3, respectively,
X(K) = xw(0) W4kx0 + xw(1) W4kx1 + xw(2) W4kx2 + xw(3) W4kx3
We have the following results:
X(0) = 0.6667
X(1) = -1.3334 + j0.6667
X(2) = 2
X(3) = 1.3334 j0.6667
f=

1
1
=
=25Hz
NT 4 . 0.01

Applying Equations (4.19), (4.22), and (4.23) leads to

A0 =

1
0
= 00
|X(0)| = 0.1667, = tanh
4
0.6667

P0 =

1
2
2 |X(0)| = 0.02778
4

A1 =

1
0.6667
= 153.440
|X(1)| = 0.3727, = tanh
4
1.3334

P1 =

1
2
2 |X(1)| = 0.1389
4

A2=

1
0
|X(2)| = 0.5, = tanh = 00
4
2

P2 =

1
2
2 |X(2)| = 0.25
4

A3=

1
0.6667
= -26.560
|X(3)| = 0.3727, = tanh
4
1.3334

P3 =

1
2
2 |X(3)| = 0.1389
4

b. Hamming Windows
N = 4, maka:
Whm(0) = 0.54 0.46 cos

2x 0
=0.08
41

Whm(1) = 0.54 0.46 cos

2x 1
=0.77
41

Whm(2) = 0.54 0.46 cos

2x 2
=0.77
41

Whm(3) = 0.54 0.46 cos

2x 3
=0.08
41

the windowed sequence is computed as

xw(0) = x(0) x Whm(0) = 4 x 0.08 = 0.32
xw(1) = x(1) x Whm(1) = -1 x 0.77 = -0.77
xw(2) = x(2) x Whm(2) = 2 x 0.77 = 1.54
xw(3) = x(3) x Whm(3) = 1 x 0.08 = 0.08
Applying DFT Equation (4.8) to xw(n) for k = 0, 1, 2, 3, respectively,
X(K) = xw(0) W4kx0 + xw(1) W4kx1 + xw(2) W4kx2 + xw(3) W4kx3
We have the following results:
X(0) = 1.17

X(1) = -1.22 + j0.85

X(2) = 2.55
X(3) = -1.22 j0.85
f=

1
1
=
=25Hz
NT 4 . 0.01

Applying Equations (4.19), (4.22), and (4.23) we achieve

A0 =

1
0
= 00
|X(0)| = 0.2925, = tanh
4
1.17

P0 =

1
2
2 |X(0)| = 0.0856
4

A1 =

1
0.85
= 145.140
|X(1)| = 3717, = tanh
4
1.22

P1 =

1
|X(1)|2 = 0.1382
42

A2=

1
0
|X(2)| = 0.6375, = tanh = 00
4
2

P2 =

1
2
2 |X(2)| = 0.4064
4

A3=

1
0.85
= 214.860
|X(3)| = 0.3717, = tanh
4
1.22

P3 =

1
|X(3)|2 = 0.1382
42

c. Hanning Windows
N = 4, maka:
Whn(0) = 0.5 0.5 cos

2x 0
=0
41

Whn(1) = 0.5 0.5 cos

2x 1
=0.75
41

Whn(2) = 0.5 0.5 cos

2x 2
=0.75
41

Whn(3) = 0.5 0.5 cos

2x 3
=0
41

the windowed sequence is computed as

xw(0) = x(0) x Whn(0) = 4 x 0 = 0
xw(1) = x(1) x Whn(1) = -1 x 0.75 = -0.75
xw(2) = x(2) x Whn(2) = 2 x 0.75 = 1.5

xw(3) = x(3) x Whn(3) = 1 x 0 = 0

Applying DFT Equation (4.8) to xw(n) for k = 0, 1, 2, 3, respectively,
X(K) = xw(0) W4kx0 + xw(1) W4kx1 + xw(2) W4kx2 + xw(3) W4kx3
We have the following results:
X(0) = 0.75
X(1) = -1.5 + j0.75
X(2) = 2.25
X(3) = -1.5 j0.75
f=

1
1
=
=25Hz
NT 4 . 0.01

Applying Equations (4.19), (4.22), and (4.23) we achieve

A0 =

1
0
= 00
|X(0)| = 0.1875, = tanh
4
0.75

P0 =

1
|X(0)|2 = 0.352
42

A1 =

1
0.75
= 153.440
|X(1)| = 0.4193, = tanh
4
1.5

P1 =

1
|X(1)|2 = 0.1758
42

A2=

1
0
= 00
|X(2)| = 0.5625, = tanh
4
2.25

P2 =

1
|X(2)|2 = 0.3164
42

A3=

1
0.75
= 206.560
|X(3)| = 0.4193, = tanh
4
1.5

P3 =

1
|X(3)|2 = 0.1758
42

12).

Penyelesaian :
a.

Pada grafik di atas menunjukkan adanya kebocoran spectral karena

amplitudonya tidak kontinyu pada domain waktu

b.

Comparison of a spectrum without using a window function and a spectrum using a

hamming window of size of 73 sample

Pada grafik di atas menunjukkan adanya kebocoran spectral karena

amplitudonya tidak kontinyu pada domain waktu