IEEE TRANSACTIONS ON COMMUNICATIONS, VOL 42, NO.

10,OCIOBER 1
W

A Technique for Orthogonal Frequency Division

Multiplexing Frequency Offset Correction
Pad H.M m e , Member, IEEE

is beiig suggested as an efficient modulation for applications

ranging from modems [4], to digital audio broadcast [5].
Oneof theprincipaladvantages of OFDM isitsutility
for transmission at very nearly optimum performance in un-
equalized channels and in multipath channels. As described in
[3]+5], htersymbol inkxfereace,(ISI)and intercarrig interfer-
ence (KT) cxtn be entitely eliminated by the s
of inserting between symbok a d l time in* km@a as
aguardinterval. The lengthof the guard inbw+,.i$~&de
equal to or greater than the time spread of the c&m+. E,into
this guard interval, the symbol signal wavefom, >'@mtmled
periodically, orthogonality of the carriers is m$nfained;qr
the symbol period,thvs eliminatingICI. ~lscx~.&ke sBcGbg9tye
symbols do not overlap because of the guard interval, IS1 is
e l i i a t e d , too.
One of the principal disadvantages of OmlM is sensitivity to
frequency offsetin the channel.For example, the codedOFDM
system developed by CCElT (Cenhe.Commun. d'Etudes de
I. INTRODUCTION Telediffusion et Telecommuications) for digital sound broad-
casting to mobile receivers incorporates an AFC (automatic
T HE TECHNIQUE demibed in this paper has been de-
veloped to correct frequency offset errors in digital com- frequency control) loop in the receiver to reduce frequency
munications systems employing orthogonal frequency divisionoffsetcausedbytuning oscillatorinaccuraciesanddoppler
multiplexing (OFDM)as the method of modulation. The aim shift [6].
of the paper is two€ol&to show theeffect offset errors have on There are two deleterious effects caused by frequency offset;
the signal-to-noise ratio of the OFDM carriers and to present one is thereductionofsignalamplitude in theoutput of
analgorithmto eswate offset so that it mayberemoved the filters matched to each of the carriers and the second is
prior to demodulitbn. introduction of IC1 from the other carriers which are now no
OFDM is a bandwidth efficient signalling scheme for dig- longer orthogonal to the filter. Because,OFDM, in the carriers
ital communicationsthatwasfirstproposedbyChang [l]. are inherently closely spaced in fkquency compared to the
The main difference between frequency division multiplexing channelbandwidth,thetolerablefrequency offset becomes
(FDM)andOFDM, is that in OFDM thespectrum of the a very small fraction of the chaonel bandwidth. Maintaining
individual carriers mutually overlap, giving thereforean opti- sufficient open loop frequency accuracy can become difficult
mum spectrum efficiency (asymptotically Q b/Hz for 2Q-ary in links, such assatellite links with multiple frequency transla-
modulation of each carrier). Nevertheless, theOFDM carriers tions or,as mentioned previously, in mobile digitalradio links
exhibit orthogonality on asymbol interval if synthesized such that may also introduce significant Doppler shift. The effects
that they are spaced infrequencyexactlyatthereciprocal of frequency offset are presented in Section II.
of thesymbol interval. Fortunately, this synthesiscan be In Section 111, we present an algorithmto estimate frequency
accomplishedperfectly, in principle,utilizingthediscrete offset from the demodulated data signals in the receiver. The
Fourier transform (dft)as first described by Darlington [2] and algorithmextendsto OFDM, withimportantdifferences,a
later, for data modems, by Weinstein and Ebert [3]. With the method described by Simon and Divsalar[7] for single carrier
recent evolution of integrated circuit digital signal processing MPSK.Thetechniqueinvolvesrepetitionof a data symbol
(dsp) chips, OFDM has become practical to implement and and comparison of the phases of each of the carriers between
Paper approved by M. J. Joindot, the Editor for Radio Communications of the successive symbols. Since the modulation phase valuesare
the IEEECommunications Society. Maauscriptreceived November 27, 1991; not changed, the phase shift of each of the carriers between
revised June 29,1992. successive repeated symbols is due to the frequency offset.
The. author is with the Electrical and Computer Engineering Department,
Naval Postgmduate School, M m t m y , CA 93943, USA. The frequency offsetis estimated using a maximum likelihood
EEE Log Number 9401947. estimate (MLE) algorithm. Performance of the algorithm as
0090-6778/94$04.00Q 1994 IEEE
a function &/No (individual OFDM carrier energy to one- synchronization at thisstage of thereceiver)leading to a
sided spectral density of additive white Gaussian noise) and receivedsequence as givenby (4). It is assumedthatthe
freq&cy offset is i u c l u ~in Section III. impulseresponseof the channeldoesnotchange(much)
In the eventthatthefrequencyoffsetexceeds f l / 2 the duringthesymbolplus gnard interval (this correspondsto
intercarrier spacing, the maximum limits of the algorithm, a "slow-fading" in a radio frequency channel).
strategy is required for initial acquisition. One such strategy The insertion of guard intervals renders the receivedcarriers
is described in Section N. orthogonalonthe N pointsymbolinterval.However,the
demodulation process, which is implemented with a dft (the
n. omhi'hWNSLAnON IN A dft is equivalent to matched filter reception in the absence of
CHANNEL WITH FREQUENCY OFFSET frequency offset) is affected by frequency offset. That is,
An OFDM transmission symbol is givenby the N point N-1
complex modulation sequence
n=O
X
xn.= (1/N) XkefLrinklN; the kth element of the dft sequence, consists of three com-
k=-K ponents;
TZ = 0, 1, 2 , . ...N - 1; N 3 2K + 1. (1)
Yk = (x&k){(sinm)/Nsin ( m / N ) }
+
It consists of 2K 1 complex sinusoids which have been .&re(N-l)/N + + wk. (7)
modulated with 2K + 1 complex modulation values {xk}. Ik

we &
te that the hdividual sinusoids are orthogonal on the The first component is the modulation value x k modified
symbol interval, that is
by the channel transfer function. This component experiences
N-1 an amplitude reduction and phase shift due to the frequency
c % k & = (1/N)Ixk12skl (2) offset. Since Nis always much greater thanXE,N sin (?rc/N)
n=O may be replaced by r e .
where Xnk = (l/N)XkeZrink/N. The second term is the IC1 caused by the frequency offset
We also note that the N point discrete Fourier transform and is given by
(dft). of (1) is ,the N point sequence K
, a

4= (XI&){(sin?re)/(Nsin(x(l- k E)/N))) +
1s-K
l#k
. ej7re(N-l)/Ne-jr(l-k)/N . (8)

= {XO,
XI,.-.XK,
0, 0 , . ..o, 0,x - K . ..x-2,X I } In order to evaluate the statistical properties of the ICI, some
(3) furtherassumptions m necessary.Specifically, it will be
a S S U l l l e d that E[Xk] = 0 and E[XkX;] = IXl26lk, that is,
ofmodulationvalues.Equation(1) is theinversediscrete the modulation values have, zero mean and are uncorrelated.
Fourier transform (IDFl') of (3) anddefinesapractical Wi th this provision E[&]= 0, and
modulation-carrier synthesis technique for generating OFDM
K
with perfect o~ogonality.
After passingthrough a bandpasschannel,thecomplex
e
ndop e of the & vd sequence can be expressed as
e l#k

/ {Nsin(?r(l- L
I=:, 1
.{sinac}2 -I-~ ) / N ) } z . (9)
y,, = <1/N)
XJ$ke2xjn(k+c)lN
+pun;
The average channel gain, E{IHi12} = lHI2, is constant so it
n=o, 1 , 2,..., N - 1 (4) can be separated from the sum and (9) becomes
where & is the transfer function of the channelatthe E[1&12]= IX121H12(sin.lrc)2
fresueacy of the kth m e r , t is the relative frequency offset K-k
of the channel (the ratio of the actual frequency offset to . l / { N s i n ( r ( p + ~ ) / ~ ) } ~(10)
.
the i&xmnier spin&, and w,, is the complex envelope of p=-K-k
additive white Gaussian noise (AWGN). Let the actual symbol ##O

transmitted be the N + Ns point sequence The sum in (10) can be bounded for E = 0. It consists of 2K
{ x N - N ~ , . . .,XN--2, XN-I, ~ ~ 5 1 . -,, ~. N - I } (5) positive terms. The intmval of the sum is contained within the
longer interval -2K 5 p <_ 2K,ita loaation dependent on k.
with Ng greater than orequal to thetimespread ofthe Recall that2K 5 N - 1. Also note-thefollowing; the argument
channel. The N, point precursor signal allows the received of the sum is periodic with period N , it is an ev& function of
symbol sequence to. reach steady state by n = 0 (we assume p, and itis even about p = N / 2 . Thus the 2K terms of the sum

1
2910 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 4 2 NO. 10, OCIDBER 1994

m. WWCY OFFSETE ~ T I M A ~ O N
If an OFDM transmission symbol is repeated, one receives,
ih the absence of noise, the 2N point sequence
r~ 1

= 0, 1,.. . , 2 N - 1. (16)
The kth element of the N point dft of the first N points of
(16) is
N-1

-5
aI am 0.1 0.15 az nz a3 0.35 a4 0.45 05
n=O
Fig. 1. S N R versus relative frequency offset.
andthekthelement of the dft of thesecond half of the
sequence is
N terms in the intervals -N/2 5 p 5 -1
are a subset of the 2N-I
and 1 5 p 5 N/2 for every k. Consequently,
K-k NI2 n=N
N -1
l/(Nsin?rp/N)' < 2 E l / ( N s i n ( ~ p / N ) ) ' . (11)
p=--K--k p=l = C T ~ + N ~ - ' ~ ' ~k ~=/0,~ 1,.
; .. ,N - 1.(18)
PZO
n=O
) ~ Ipl 5 N/2. There-
Observe that (sin~pPjN)"2 ( ~ P / N for
fore,

N f2 103
< 2c1/(2p)2 < - c l / p 2 = 7r2/12 = 0.882 (12) Observe that between the first and second DFT's, both the
p=1 Zp=1 IC1andthesignal are altered inexactlythesameway,by
upper boundsthesum for small E. Numerically,wehave a phase shift proportional to frequency offset. Therefore, if
dekrmined that the sum in (10) isboundedby 0.5947 for offset E is estimated using observations (20) it is possible to
E < 0.5 so that obtain accurate estimates even when the offsetis too large for
satisfactory data demodulation.
E[11k1*] ; 5 0.5
5 0 . 5 9 4 7 1 X 1 2 1 H 1 2 ( ~ i n ~ ~ ) 2161 (13) It is shown in the Appendix that the maximum likelihood
upper bounds the variance of the IC1 for values of frequency estimate (MLE)of E is given by
offset up to plus or minus one half the carrier spacing. i= (1/27r) tan-1
Equation (13) may be used to give a lower bound for the
S N R at the outputof the dft for the OFDM carriers in a channel
with AWGN and frequency offset. Thus,
SNR 2 l X l " l H 1 2 { S i n ? r E / ( ~ E ) ) 2 / This is an intuitively satisfying result since,in the absence of
E each k . Fig. 2 shows
{0.5947)X121H12'(Sin?r~)2 + E [ l w k 1 2 ] } . (14) noise, the angle of Y 2 k Y i k is ~ A for
simulation results for the estimate of E obtained using (21)
It is easilyestablishedthat ~ X ~ 2 ~ H ~ 2 / E=[ /EJN,, W ~ ~ 2 versus
] E for values of EJN, corresponding to 17 and 5 dB.

where E, is the averaged received energy ofthe individual

carriers and N0/2 is the power spectral density of the AWGN A. Statistical Properties of the Estimate
in the bandpass transmission channel. Therefore, (14) may be The conditional mean and variance of E^ given E and { R k }
more conveniently expressed as can be approximated as follows. Consider the complex prod-
ucts Y2kY:k from which we estimate e. For a given E, subtract
Sm L { W N o ) { s i n d ( ? r ~ ) I 2 /
the corresponding phase, 2ae, from each p&ct to obtain the
{1+0.5947(E,/N,)(sin?r~)~} (15)
tangent of the phase e m
with equality atE = 0. Equation (15) is plotted in Fig. 1 versus
frequency offset E for values of EJN, equal to 11, 17, 23, tan [27r(i - E ) ] =
and 29 dB. Simulation results for N = 256, K = 96, and 8
{(
k$~[yZky~ke-2T3'

FSK modulation in a flat channel are included for validation.

It can be seen that the bound of (15) is quite tight for small
vdues of E but is &out 3 dB too low at E = 0.5.
 lW
Oo OM a1 au a2 a25 a3 au a4 0.6 a5 20 25 30 35 40 45 50 55 60
Fig. & MLE relative frequency offset estimate versus relative frequency Fig. 3. MLE relative frequency offset estimate error standard deviation
offset versus E,IN,.

Por 12 - E ( << 1/27r,the tangent can be approximated by its

e = 0 and 0.45. The theoretical standard deviation from(25) is
argument so that plotted for comparison. We conclude that (21) will give very
accurate estimates ofthe relative frequency offset E. Under
i - e M (1/2r) normal conditions for communications signalling, the accuracy
is sufficient to correct E to well within tolerances (see (14)
and Fig. 1) for negligible losses due any residual error in the
offset estimate.
The l i i t s of accurate estimation by (21) are I E ~ 50.5, that
is, f1/2the intercarrier spacing. As E +OS, E may, due to
noise and the discontinuity of the arctangent, jump to -0.5.
(23) When this happens the estimate is no longer unbiased and,
in practice, it becomesuseless. Thus, for frequencyoffsets
At high signal-tcmoise ratios, a condition compatible with suc-
cessful communikations signaling, (23) may be approximated exceeding one half the carrier spacing, an initial acquisition
bY strategy must be prescribed. One such strategy is discussed in
Section IV.
z - E w (1/274
B. Frequency ofsset Estimution in a Multipath Channel
It is evident from (25) and (26) that the mean and variance
/ K \ \ of the offset estimate do not depend on the actual received
frequency coefficients { R k } . Furthermore, if the symbol pair
has been received through an unknown multipath channel, and
as described in Section I it has been preceded by a periodic
from which we lind that precursor of length NB2 N , the time spread of the channel
E [ i - t I E , {Rk}] = 0. (25) then the carriers remain at their Steady state values throughout
the duration of both symbols because the modulation values
Therefore, for small errors, the estimate is conditionally un- are repeated. Thus, as no guard interval is required between
biased. the symbol pair, the algorithm of (21) can be used without
conditional varianceof the estimateis easily determined modification.
for (24). Fig. 4 shows six amplitude responses of a channel with five
multipaths whose arrival times have been uniformly distributed
I € 9 {&)I = 1/{(2~)2(Es/No>) (26) over an interval T, = T/16.The paths have equal weight and
where random phases so that the overall channel exhibits frequency
N-1 selective Rayleigh fadingas is evident from the figure. Fig. 5
showsestimatesof E from (21) with PIg = N/16 for the
n=O same conditions as Fig. 2. It can be seen that the estimate is
unaffected by the multipath.
is the total symbol energy. Since the total energy is the sum
+
of the energies of the 2K 1 carriers, the error variance
IV. ACQUISITTON
of the offset estimate will in practice. be verylow.Fig. 3
shows the sample standard deviation of theerror in the relative In the event that the frequency offset is greater than f 1 / 2
fresuency offset estimate for 100 simulationtrials of (21) the carrier spacing, a strategy for initial acquisition to bring
+
versus Es/No for 2K 1 =193 carriers and for offsets of the offset withinthe limits of the algorithmmust be developed.
 2912 II3 E E TRANSACTIONS ON COERMUNICATONS, VOL. 42, NO. 14 OCTfBJ3R 1994
I

10, 1 repeated symbols by a repetition of the first full length data

symboi or by the use of an AFC loop as shown in [6].
To illustrate, consider the following example for a digital
audio broadcasting service. Assume an intercarrier spacing for
the data stream of 1 kHz, and a frequency offset uncertainty
in thesystemdominatedbythelong term accuracy of the
oscillators in the receiver that heterodyne the received signal to

Ii \I IF and quadrature demodulate to obtain the complex envelope.

Assume VHF radio transmission at 150 MHz andoverall
oscillator uncertainty (long term stability) of 1 part in lo5.
Thus,,,,S =1500 Hz and Afinit;al must be greater than3000
I , I Hz. At regular intervals in the datastream of 1 ms (plus guard
-100 -sa 0 so 100 150
interval) symbols insert a short symbol of length 250 p s (4
Fig.4. M I I ~ I ~ @ I . ' chaunel. lcHz carrier spacing) and repeat it once. From this xpeated
shortened symbol estimateE. Assume E,/N, is 11 dB for the
. 0 3
data symbols and that there are 200 carriers, so that E./N, is
. . 34 dB. The shortened symbols have 1/4thenergy so E,/N,, is
a45 -
only 28 dB for the initial estimateof E . From (25) we find that
a4 -
DEinitial = 0.0063 SO that Dcdata = 0.025. From (15) we S e e
0.3- that residual offseterrors of this magnitude causeIC1 resulting
in a signal-to-interference ratio of24.3 dB in the data symbols
for only 0.2 dB loss in overall S N R at EJN, of 11 dB.
a2. &fNo=l7 dB This exampleillustratesasituation for whichthe initial
an - offset acquisition estimate with shortened data symbols is of
0=5 dB
at sufficient accuracy that refinementof the estimate with longer
/:i
oo
aos -
.'
,....,
am
a2 azs
'
,.,'

,--
.

..
a3 as a4 0.4s
a1 ais as
symbols is not necessary.

F 5 MLE dative fresuency offset estimate versus relative frequency

0k.ba multipath c-1.
v. DISCUSSIONAND CONCLUSIONS
We have seen that, as expected, frequency offset in OFDM
causes serious loss of SNR of the dft outputs due primarily to
We envision that,if continuous, theOFDM symbol stream will ICI. A lower bound for S N R has been derived and simulation
be punctuated at appropriate intervals withqpeated symbols. results show that the bound is quite accurate for small offsets,
A continuous symbol stream occurs in applications such as but about 3 dB too pessimistic as the offset approaches 112
di&d audio broadcast[5]. A second possibility is thatOFDM the carrier spacing.
modulation is used in session oriented digital data or voice
communications such as digital radio [8]. Here, we envision
Analgorithm for maximumlikelihoodestimate
of frequency offset using the dft values of a repeated
that the session hitiation interval will include one or more symbol has been presented. It has been shown that for small
(92
lqeated symbols. error inthe estimate, the estimate is conditionally unbiased
%e basic strategy for initial frequency offset acquisition,
I

and is consistent in the sense that the variance is inversely

in &her case, is to shorten the dft's and use larger carrier proportional to the number of carriers in the OFDM signal.
spacings such that the phase shift does hot exceed f ~The
. Furthermore,boththesignalvaluesand the IC1 contribute
i i q k n c y offset in Hz is S = t / T = eAf Where Af is the coherently to the estimate so that it is possible to obtain very
intercarrier spacing and T is the symbol interval. Let us assumeaccurate estimates even when the offset is too great, that is
thattheinitialfrequencyoffset is nogreaterthanThen there is too much ICI, to demodulate the data values. Since
the estimation error depends onlyon total symbol energy, the
algorithm works equally well in multipath spread channels.
However, it is required that the frequency offset as well as
determinesthe -um
initial canier spacing,and corre- the channel impulse response be constant for a period of two
spoutling dft lengths. If the average power of the shortened symbols.
symbols.is keptthe m e , the varianceof the estimate of einitial Theaccuracyrequiredoffrequencyoffsetcorrectionde-
will 'be larger than for the longer data symbols since there pends on how much residual offset can be tolerated. Offset
is less symbol energy. Also, the offset estimate error for the inducedIC1 can be treated quite satisfactorily as additional
shortened symbols, since it estimates the fraction of carrier AWGN since its s o m e is the multiplicity of other OFDM
spacing,corresponds to aproportionatelylargerfractional carriers thatare zero mean and uncomlated random processes.
offset f a the longer data symbols. However, theMLE estimate Note that the SNR defined in Section II [see (14)and (15)]
is .sa accurate that in practice the initial estimate still may is just EJN, in the absence of offset. Thus we may interpret
be adequate. If not, it is refined by following the shortened (15) as theeffective E,/N, of thecarriers, or if divided
Moos6: OPDM FREQUF.NCY OFFSET CORREcIlON 2913
b
\
by the number of bits encoded in each of the carriers, their To find the conditional density function in (A.8), note that
effectiire &/No.ReSUired Eb/N, O f COurSe depends u p the
moddation constellation. the fading statistics of the chaanel Yz = (Y1-W , ) H ( @+) wz 64.9)
theforwarderrorcontrolcoding, if any,employed m the so that
OPDM system and the desired BER (see, for example, [ 5 ,
Figs. 11-13]). +
Yz = Y l H ( 8 ) Wz - W l H ( 8 ) . (A.10)
The aquisiion range of the algorithm presented here is
f1/2the intercarrier spacing of the repeated symbol, It is If W1 and W2 are Gaussian, zero mean whiterandom
independent of the modulation constellations chosen for the vectors with variancea’,then the conditional density function
carriers and whether the symbols are coherently or differen- in (A.6) is multivariateGaussianwith mean valuevector
tially encoded.The AFC loop shown in [6] does not require a Y l H ( 6 ) and 2M x 2M covariance matrix
repeated symbol. However its acquisition range isonly f1/2m K = E[(W2- W1H(@))t(W2 - wlH(e))] =221.
of the intercarrier spacingfor m-ary PSK. The initial frequency (A.11)
offset at the timeof the initiationof the communicationsession We note that K is independent of 8, therefore,
may be greater than 1/2 theintercarrierspacingandthus !
even outside the range of the MLE algorithm. In this event, 6 = r n ~ [ f ( 18,
~ z Y , ) ]= %[~(8)1 (A.12) I

a strategy is required for initial acquisition. We propose to

use a pair of shortened data symbols whose carrier spacing with
is sufficiently large to insure that the algorithm will operate J ( 6 ) = (Yz- YlH(Q))(Yz - YlH(8))’. (A.13)
within its range. Dueto the low varianceof the initial estimate,
further refinement will normally not be required. It may be Using the fact that
advantageous to use shortened repeated symbols for tracking
offsetvariationstoo,insteadof an AFC loop, because this +
H(O)[dH(Q)/dO]’ [ d H ( @ ) / d @ ] H t ( 6=) 0 (A.14)
reduces the time during which the channel must be stable. we canfind that
APPENDIX d J ( Q ) / d Q = -Y2[dH(Q)/dQItYE- Yl(dH(@)/d@]Yi.
”UMLIKELIHOOD ~ ” I M A T EOF DIFFERENTJAL PHASE (A.15)
1

Let ibf complex values (Zk}be representedbyalength Using (A.4), it follows directly that (A.15) is identically zero
2M row vector when 0 = Q such that
I
= [ Z l R Z2R ’ . .Z M R ZlI &I ’* ’Z M I ]
+
sin ( ~ ) [ Y Z R Y : R y z r % I l = cos ( ~ ) [ Y ~ ~ Y-Ey zRR Y , t l l .
(A.16)
= [ZR 211. (A. 1) Therefore,
Consider the random vectors
Y1=R1+W1 (-4.2)
Y2 = RlH(6) + w2 (A.3)
L

where (A. 17)

qe)= E g], C = c o s ( e ) ~& s = sin(e)I is the maximum likelihood estimate (MLE) of 0.

(A.4)
is a 2M X 2M rotationmatrix.Themaximumlikelihood
estimate of the parameter 8, given the observations Y1 and
Y2 (see,for example, Sage and Melsa,[9, p. 1961) is the value
of 8 that maximizes the conditional joint density function of
the observations. That is
8 = m$f(Y1, Yz I e)]
ACKNOWLEDGMENT
The authorwould l i e to thank Dr. S . Pupolin of University
of Padova for his helpful suggestions to the original manu-
script. The authoris also indebted to theanonymous reviewers
whose constructive remarks have improved thequality of this
(A.5) Paper-

which can be written as REFERENCES

e = me!$f(Y2 18, Y M Y l I @)I. (A.6) [I] R. W. Chang, “Synthesis of band-limited orthogonal signals for multi-
channel data transmission,” Bell Sysr. Tech J., vol. 45, pp. 1775-1796,
Dec. 1966.
But 9 gives no information about Y 1 ,that is 121 S. Darling, “On digital single-sideband modulators,”IEEE Trans. Circuit
Theory, vol. CT-17, pp. 409414, Aug. 1970.
W l I Q) = f w d (A.7) [3] S . B. Weinstein and P. M. Fibert, D ’‘ ata transmission by frequency-
division multiplexing using the discrete Fourier transform,” IEEE Trans.
so that Comnutn. Technol., vol. COM-19, pp. 628-634, Oct. 1971.
[4] J. A. C.Bingham, ‘“ulticarriw modulation for data transmission: An
6 = m z [ f ( Y 2I 8,Y l ) ] . (A.8) idea whose time has come,” IEEE Cornnurn. Mug., vol. 28, pp. 17-25,
Mar. 1990.
 - _
~- _ ~- ~-

I
pJY---.---~L- ~ ~ ~ .~ ~ ~~ ~ ~ ~ 1

2914 IEEE TRANSACTIONS ON COMMUNICATIONS. VOL. 42. NO. IO, O C m B W 1994

M. Nard and R Halbert. “Principlesof modulationchannel

and coding ppulM.Moose(M’79)wasbornin0maha,NE,
for digital broadcasting for mobile receivers,” EBU Rev., no. 224. pp. on July 22. 1938. He received the B.S., M.S., and
3-25, Aug. 1985’. Ph.D. d ew in elecnical engineering in 1960,
[6] B. LeFlocb, R. Halbert-Lassalle, and D. Ca~telain.“Digital sound 1966, and 1970, respectively, from thc University
brosdcasting to mbde receivers,” IEEE Ifm.C o m m Ekc.. voL 35, of Washington. Seattle.
no. 3, pp. 493-503, Aug. 1989. Since 1977 he has been with the Department Of
[ 7 l M. K S i and D. Divsalar, “Doppler-ted differential detection Electrical andComputer Engineering, Naval Past-
of MPSIC’IEEE Tram. Commun.. vol. 37, no. 2, pp. 99-109. Feb. 1989. ,~ graduate School, Monterey, CA, wberehe is an
[8] D.C. Cox, W.S. G80dand H.Sherry, “Low-power digital 1adi9asa Associstt Rofess~rof Blectrical and COmpMer En-
ubiquhm subscriber loop,” IEEE Commm Mag., vol. 29, pp. 92-95, gineoring. His research interests 82e in digital com-
Mar. 1991. munications. He is a -founder of and consultant
[9] A. P. Sage and 1. L. Melsa, Estimation Theory with Applicarions to fir M~~ c o m m u ~h.,
~igital c ~~ o m ~CA. ~ w ~ ,
Commwicolionr and Control. New York:McGraw-Hill, 1971.