EURASIP Journal on Wireless Communications and Networking 2004:1, 67–73c© 2004 Hindawi Publishing Corporation

Downlink Space-Frequency PreequalizationTechniques for TDD MC-CDMA MobileRadio Systems

Adao SilvaInstituto de Telecomunicacoes, Universidade de Aveiro, Campus Universitario de Santiago, 3810-193 Aveiro, PortugalEmail: asilva@av.it.pt

Atılio GameiroInstituto de Telecomunicacoes, Universidade de Aveiro, Campus Universitario de Santiago, 3810-193 Aveiro, PortugalEmail: amg@det.ua.pt

Received 31 October 2003; Revised 16 March 2004

The paper considers downlink space-frequency preequalizations techniques for time division duplex (TDD) MC-CDMA. We con-sider the use of antenna arrays at the base station (BS) and analytically derive different preequalization schemes for two differentreceiver configurations at the mobile terminal: a simple despread receiver without channel equalization and an equal-gain com-biner (EGC) conventional receiver. We show that the space-frequency preequalization approach proposed allows to format thetransmitted signals so that the multiple access interference at mobile terminals is reduced allowing to transfer the most compu-tational complexity from mobile terminal to the BS. Simulation results are carried out to demonstrate the effectiveness of theproposed preequalization schemes.

Keywords and phrases: MC-CDMA, preequalization, antenna array, TDD, downlink.

1. INTRODUCTION

The beyond 3G broadband component of wireless systemmust be able to offer bit rates of more than 2 Mbps in a ve-hicular environment and at least 10–20 Mbps in indoor andpedestrian environments [1].

It is consensual that MC-CDMA is one of the mostpromising multiple-access schemes for achieving such highdata rates [2]. This scheme combines efficiently orthogonalfrequency division multiplex (OFDM) and CDMA. There-fore, MC-CDMA benefits from OFDM characteristics suchas high spectral efficiency and robustness against multipathpropagation, while CDMA allows a flexible multiple accesswith good interference properties for cellular environments[3].

Recent publications have shown that MC-CDMA is par-ticularly advantageous for the downlink, that is, from BSto mobile terminal (MT) [4]. However, the user capacityof MC-CDMA system is essentially limited by the multiple-access interference (MAI) provoked by the loss of code or-thogonality among the users in multipath propagation. Usu-ally, in conventional MC-CDMA downlink, the MAI is mit-igated by frequency domain equalization techniques at re-

ceiver side. Since low complexity is required at MTs, onlysimple detection techniques can be implemented, limitingthe MAI cancellation capability. Considering time divisionduplex (TDD), another solution consists in performing pree-qualization at the transmitter side using the TDD channelreciprocity between alternative uplink and downlink trans-mission period [5, 6, 7]. The knowledge of the channel stateinformation (CSI) from uplink can be used to improve theperformance in downlink. The crucial assumption is thatchannel dynamics are sufficiently slow so that the multipathprofile remains essentially constant over the block of trans-mitted symbols. Normally, this principle is valid for indoorand pedestrian environments, that is, in low-mobility sce-narios. The aim of this solution is to allow the use of simplelow-cost, low-consuming MT.

It is well known that the use of antenna arrays increasesthe system capacity by reducing the effect of frequency se-lective fading and improving the spectral efficiency, withoutadditional frequency spectrum [8]. It has been show [9] thatthe combination of antenna array with MC-CDMA system isvery advantageous in cellular communications.

This paper proposes a downlink TDD downlink MC-CDMA space-frequency preequalization algorithm designed

68 EURASIP Journal on Wireless Communications and Networking

User 1 QPSKmod d1

c1,0, . . . , c1,L−1

d1,0× L w1,0

0...

P − 1L×

d1,p−1

c1,0, . . . , c1,L−1

w1,p−1

L... M

+

L

L

... M

+

L

...IFFT

+GP

Antenna 1

Hg,1

FFT−

GP

0

cg,0, . . . , cg,L−1

×

+dg...

L− 1×

cg,0, . . . , cg,L−1

Mobile terminal (a)... CSI

...

User k QPSKmod dk

ck,0, . . . , ck,L−1

dk,0 × L wk,00

...

P − 1

Base station

L×dk,p−1

ck,0, . . . , ck,L−1

wk,p−1

L... M

+L

L

... M

+L

...IFFT

+GP

Antenna M

Hg,m

FFT−

GP

0...

L− 1

wr,0

×

cg,0

×

+dg

× ×

...

wr,L−1 cg,L−1

Mobile terminal (b)

Figure 1: Transmitters and receivers schemes for downlink MC-CDMA.

for two different types of receivers: equal-gain combiner(EGC) conventional equalizer and a very simple receiverwithout channel equalization. Both algorithms operate in thefrequency domain and optimization is done in frequency forthe single-antenna case, and jointly in space and frequencywhen considering an antenna array. Moreover, these algo-rithms are designed using as criterion the minimization ofthe transmitted power at the BS, which is a very importantissue for most preequalization algorithms.

The paper is organized as follows. In Section 2, wepresent the proposed downlink MC-CDMA system. InSection 3, we analytically derive the preequalization algo-rithms for the two receiver configurations which we call con-strained zero forcing (CZF) and CZF-EGC, respectively. InSection 4, we present some simulation results obtained withthe CZF and CZF-EGC preequalization techniques in twodifferent scenarios, beamforming and diversity, and compareboth preequalization schemes against conventional equalizertechniques such as MRC, EGC, and MMSE. Finally, the mainconclusions are pointed out in Section 5.

2. SYSTEM MODEL

Figure 1 shows the proposed downlink MC-CDMA trans-mitters and receivers. As presented in Figure 1, for each userk, a complex QPSK data symbol dk (k = 1, . . . ,K) is con-verted from serial to parallel to produce p symbols, dk,p

(k = 1, . . . ,K and p = 0, . . . ,P − 1), where P denotes thenumber of data symbols transmitted per OFDM symbol.The data symbols are spread into L chips using the orthog-onal Walsh-Hadamard code set and scrambled by a pseudo-random code. We denote the code vector of user k as ck =[ck,0, . . . , ck,L−1]T , where (·)T is the transpose operator. Then,

the chips of the data symbols are copied M times in or-der to obtain L ·M versions of the original symbols whichare weighted and transmitted over M antenna branches. TheLM chips for user k and symbol p are weighted by a vector

wk,p =[

wTk,p,1 wT

k,p,1 · · · wTk,p,M−1

]Twhere wk,p,m of size L

contains the set of coefficients that weight the chips that goto antenna m, and thus wk,p is of size LM. These weights arecalculated using the CSI according to the criteria presented inSections 3.1 and 3.2. After that, the signals of all users on eachsubcarrier and antenna branch are added to form the mul-tiuser transmitted signal. Finally, a guard period (GP) longerthan the channel multipath spread is inserted in the trans-mitted signal, on each antenna, to avoid intersymbol inter-ference (ISI).

The transmitted signal, in frequency domain, for ageneric data symbol p is given by

yp =K∑k=1

dk,p�ck ◦wk,p, (1)

where�ck =

[cTk , . . . , cTk

]Tis a column vector of size of L ·M

that represents the spreading operation and consists of Mrepetitions of the code vector for user k since the same codeis used for all antenna branches, and (◦) means an element-wise vector product. The vector signal yp of length L ·M ismapped to the antenna branch so that the first L elementsare transmitted over the L subcarriers of the OFDM modu-lation on the first antenna branch, the second L elements tothe second branch, and so on.

The input signal at the generic mobile g, for symbol p, isobtained multiplying (1) for the channel frequency responsefrom the BS to the MT of the desired user and adding AWGN

Downlink Preequalization Techniques for TDD MC-CDMA Systems 69

noise:

xg,p =M∑

m=1

K∑k=1

dk,pck ◦wk,p,m ◦ hg,p,m + ng , (2)

where hg,p,m of size L × 1 is the channel frequency responsebetween antenna m and MT.

At the receiver side we propose two different receivers:receiver (a) which is composed just by a single antenna, anFFT, despreading and descrambling operations, that is, we donot perform channel equalization, and a conventional EGCsingle user receiver (b). For this latter case the weights aregiven by

wr = h∗

|h| , (3)

that is, we just perform phase equalization.For receiver (a), the decision variable at the input of the

QPSK demodulator is, for the desired user g and symbol p,given by

dg,p = dg,p ·cg M∑

m=1

wg,p,m ◦ hg,p,m

cHg

︸ ︷︷ ︸desired signal

+K∑

k=1, k �=gdk,p · ck

M∑m=1

(wk,p,m ◦ hg,p,m

)cHg

︸ ︷︷ ︸MAI

+ nr︸︷︷︸noise

.

(4)

For receiver (b), the decision variable expression is verysimilar to (4), but in this case the vector hg,p,m is replacedby

zg,p,m =(∣∣hg,p,m

∣∣2+ hg,p,m

∑Mi=1, i �=m h∗g,p,l

)∑M

m=1

∣∣hg,p,m∣∣ , (5)

where (·)∗ denotes the complex conjugate.The vector nr represents the residual noise samples ofML

subcarriers. The signal of (4) involves the three terms: the de-sired signal, the MAI caused by the loss of code orthogonalityamong the users, and the residual noise after despreading.

3. TRANSMIT PREEQUALIZATION SCHEMES

In this section we analytically derive a space-frequency pree-qualization algorithm for the two receiver configurations: (a)and (b). In the latter case, the weights are computed takinginto account that at the receiver we have the EGC combiner.However, we use the same criterion, zero forcing, in bothpreequalization schemes.

The use of preequalization algorithms has two main ad-vantages: reducing the MAI at MTs by preformatting the sig-nal so that the received signal at the decision point is free

from interferences and allowing moving the most computa-tional burden from MT to BS, keeping the MT at a low com-plexity level. When we use an antenna array at the BS, thepreequalization can be done in both dimensions, space andfrequency. We propose to jointly optimize the user separa-tion in space and frequency by the use of criteria based onthe decision variable after despreading at the MT. This op-timization task is performed taking into account the powerminimization at the transmitter side.

3.1. CZF preequalization algorithm

The CZF preequalization algorithm is based on a zero-forcing criterion, since we put a zero in the MAI term. Thisalgorithm is designed in order to remove the MAI term of(4) at all MTs at same time. Furthermore, it takes into ac-count the transmitted power at BS, the reason we call thisalgorithm the CZF.

Applying the zero-forcing criterion to (4), we obtain thefollowing conditions:

cg ◦ M∑

m=1

wg,p,m ◦ hg,p,m

cHg = 1,

K∑k=1, k �=g

ck ◦ M∑

m=1

wk,p,m ◦ hg,p,m

cHg = 0,

(6)

ensuring that each user receives a signal that after despread-ing is free of MAI. The first term of the right-hand side of(4) is the desired signal, and has been made, for normaliza-tion purposes, equal to 1, while the second term representsthe interference caused by other K − 1 users, and accordingto the criterion used should be equal to 0.

The interference that the signal of a given user g producesat another MT k is obtained for a generic data symbol ac-cording to (4):

MAI(g −→ k) = cg

M∑

m=1

wg,p,m ◦ hk,p,m

cHk

= vk,g ◦wg,p

(7)

with vk,g = �cg ◦

[hk,p,1 hk,p,2 · · · hk,p,M

] ◦ �cHk .The weight vector for user g is then obtained by con-

straining the desired signal part of its own decision variableto 1, while cancelling its MAI contribution all other MTs atsame time. This leads to the following set of conditions:

cg ◦ M∑

m=1

wg,p,m ◦ hg,p,m

cHg = 1,

cg ◦ M∑

m=1

wg,p,m ◦ hk,p,m

cHk = 0 ∀k �= g.

(8)

Therefore, to compute the weights for user g, we haveto solve a linear system of K EQUATIONS (constraints) and

70 EURASIP Journal on Wireless Communications and Networking

L ·M variables (degrees of freedom) given by

Ag,pwg,p = b, (9)

where Ag,p is a matrix of size K ×ML, given by

Ag,p =

[hg,p,1 hg,p,2 . . . hg,p,M

]vH0,g

...vHg−1,g

vHg+1,g...

vHK−1,g

, b =

10...0

. (10)

As pointed above the prefiltering algorithms should takeinto account the minimization of the transmitted power.Therefore, the transmitted power must be minimized un-der K above constraints. When the number of constraintsequals the number of degrees of freedom, a single solutionexists provided there are no singularities. If however we havemore degrees of freedom than constraints (ML > K), thensignal design can be done to optimize some cost function,normally, the total transmitted power. The optimization willbe more effective the higherML−K is. This optimization canbe solved with the Lagrange multipliers method [10].

The transmitter power pt is given by,

pt = wHg,pwg,p (11)

and consequently we need to minimize the following costfunction with Lagrange multiplier α,

j = wHg,pwg,p − αAg,pwg,p, (12)

computing the gradient vector of j and equating to zero, weget

∇wj = 2 ·wHg,p − αAg,p = 0 (13)

thus the vector weight is given by

wg,p = 12AHg,pα

H , (14)

and then using this result in (9), we get

αH = 2 ·(Ag,pA

Hg,p

)−1 · b. (15)

Finally, replacing αH in (14), we obtain the CZF-based pre-filtering vector given by

wg,p = AHg,P

(Ag,pA

Hg,p

)−1b = AHg,pψ

−1g,pb, (16)

where ψg,p = [Ag,pAHg,p] is a square and Hermitian matrix ofsize K × K .

Observing the last equation, it easy to see that the mostcomputational intensive task, to calculate the weights, comesfrom matrix Ψg,p inversion. However, the size of this matrixis just K × K independently of the spreading factor and thenumber of antennas, which makes this algorithm very attrac-tive for practical implementations.

3.2. CZF-EGC preequalization algorithm

As referred, for this scheme we also use the zero-forcing cri-terion, but now to compute the weights, we should take intoaccount that we have the EGC combiner at receiver side, thereason we call this algorithm CZF-EGC. When we use theEGC at receiver side, we increase the complexity as comparedwith receiver (a). However, the complexity level is still per-fectly tolerable for practical implementations. EGC requiresonly knowledge about the channel phase.

The interference that the signal of a given user g producesat another MT k is obtained for a generic data symbol ac-cording to (4) and (5) by

MAI(g −→ k)

= cg

M∑

m=1

wg,p,m ◦∣∣hk,p,m

∣∣2+ hk,p,m ◦

∑Mi=1, i �=m h∗k,p,l∑M

m=1

∣∣hk,p,m∣∣

cHk

= sk,g ◦wg,p,(17)

with

sk,g

= �cg ◦

(

∣∣hk,p,1∣∣2

+ hk,p,1 ◦(

h∗k,p,2 + · · · + h∗k,p,M

)∣∣hk,p,1

∣∣ + · · · +∣∣hk,p,M

∣∣)· · ·

(∣∣hk,p,M∣∣2

+hk,p,M◦(

h∗k,p,1 +· · · + h∗k,p,M−1

)∣∣hk,p,1

∣∣ + · · · +∣∣hk,p,M

∣∣)

◦ �cHk .(18)

The weight vector for user g is also obtained by constrain-ing the desired signal part of its own decision variable toone while cancelling its MAI contribution to all other MTsat same time. This leads to the following set of conditions:

cg

M∑m=1

wg,pm◦(∣∣hg,p,m

∣∣2+ hg,p,m ◦

∑Mi=1, i �=m h∗g,p,i

)∑M

m=1

∣∣hg,p,m∣∣

cHg =1,

cg

M∑

m=1

wg,p,m ◦(∣∣hk,p,m

∣∣2+ hk,p,m ·

∑Mi=1, i �=m h∗k,p,i

)∑M

m=1

∣∣hk,p,m∣∣

cHk

= 0 ∀g �= k.(19)

As the CZF algorithm, the CZF-EGC-based pre-filteringvector is given by (16). However, the matrixAg,p is now given

Downlink Preequalization Techniques for TDD MC-CDMA Systems 71

by

Ag,p=

(∣∣hg,p,1

∣∣2+hg,p,1◦

(h∗g,p,2 +· · ·+h∗g,p,M

)∣∣hg,p,1

∣∣ + · · · +∣∣hg,p,M

∣∣)· · ·

(∣∣hg,p,M∣∣2

+hg,p,M◦(

h∗g,p,1 +· · ·+h∗g,p,M−1

)∣∣hg,p,1

∣∣ + · · · +∣∣hg,p,M

∣∣)

sH0,g...

sHg−1,g

sHg+1,g...

sHK−1,g

.

(20)

From (19) we can see that for the case M = 1 (singleantenna), we obtain an expression very similar to (8) for asingle-antenna case, given by

cg ·wg,p ◦∣∣hg,p

∣∣ · cHg = 1,

cg ·wg,p ◦∣∣hk,p

∣∣ · cHk = 0 ∀k �= g.(21)

In this case, the weights are real, because we use the mod-ulus of the channel frequency response; thus we just equalizethe amplitude at the transmitter whereas the phase is equal-ized at the receiver side.

4. NUMERICAL RESULTS

To evaluate the performance of the proposed preequaliza-tion algorithms, we used a pedestrian Rayleigh fading chan-nel, whose system parameters are derived from the EuropeanBRAN Hiperlan/2 standardization project [11]. This channelmodel has 18 taps, multipath spread of 1.76 µs and coherencebandwidth approximately equal to 637 KHz.

We extended this time model to a space model in twodifferent ways: for the diversity case, we assumed that the dis-tance between antenna elements is large enough to considerfor each user M independent channels, that is, we assumeindependent fading processes; for the beamforming case, weallocated a direction of arrival (DOA) to each path (beam-forming), with the DOAs randomly chosen within a 120◦

sector. In this latter case, the BS is equipped with a half-wavelength-spaced uniform linear array. We considered a DLsynchronized transmission using Walsh-Hadamard spread-ing sequences of length 32 scrambled by a pseudorandomcode. We used 1024 carriers, a bandwidth equal to 100 MHz,and a carrier frequency equal 5.0 GHz. The duration of theGP is 20% of the total OFDM symbol duration. The channelis considered to be constant during an OFDM symbol.

The simulations were carried out to assess the perfor-mance of the CZF and CZF-EGC algorithms in the two dif-ferent scenarios presented above, and to compare against theperformance achieved with conventional frequency equaliza-tion receivers, such as MRC, EGC, and MMSE. For a better

M = 1

M = 1M = 2

M = 8

M = 4

CZFCZF-EGCEGC

MRCMMSEAWGN

0 2 4 6 8 10 12 14 16 18 20 22

Et/No (dB)

10−4

10−3

10−2

10−1

Ave

rage

BE

R

Figure 2: Performance comparison between the CZF, CZF-EGC,and conventional receivers as function of Et/No, for diversity case.

comparison with a variable number of antennas, the resultshave been normalized, that is, for the case of multiple trans-mitting antennas, the figures do not take into account thearray gain which is 10 Log(M) in dB.

The simulation results for the diversity case are shown inFigures 2 and 3. The simulations of Figure 2 were run for anumber of users K = 32, that is, a full-load system, and themetric used is the average bit error rate (BER) as function ofEt/No, the transmitted energy (assuming a normalized chan-nel) per bit over the noise spectral density. The performanceof the CZF and CZF-EGC algorithms is illustrated for thecases of M = 1, 2, 4, and 8 transmit antennas. With a sin-gle antenna at the BS, there is no spatial separation and thepreequalization operation is done only in the frequency di-mension. As it can be seen from Figure 2, the performanceof the CZF algorithm for a single antenna is modest. At lowvalues of Et/No, the performance is even worse than with allsingle-user conventional detectors, and only for high valuesof Et/No the CZF outperforms the conventional MRC equal-izer. This occurs because for a single antenna and full loadsystem, we do not have enough degrees of freedom to mini-mize the transmitted power. The number of degrees of free-dom is equal to M · L and the number of constraints is K .Thus, for M = 1 and K = L (full-load system), the numberof degrees is equal to the number of constraints. For multi-ple antennas at BS, it is possible to optimize the preequal-ization algorithm in both dimensions, space and frequency.When we use an array of 2, 4, and 8 antennas, the perfor-mance of the CZF algorithm is much better than all single-user conventional equalizers for any Et/No value. We can seethat with 4 and 8 antennas, the performance is very closeto the one obtained with the Gaussian channel. As it can beseen from Figure 2, the performance of the CZF-EGC algo-rithm for single antenna outperforms the MRC, EGC, MMSE

72 EURASIP Journal on Wireless Communications and Networking

Et/No = 10 dB

M = 1

M = 1

M = 2

M = 4

CZFCZF-EGCEGC

MRCMMSE

1 5 10 15 20 26 32

Number of users

10−4

10−3

10−2

10−1

Ave

rage

BE

R

Figure 3: Performance comparison between the CZF, CZF-EGCand conventional receivers as function of number of users, for di-versity case.

conventional equalizers, and the CZF. This occurs because,as can be seen from (21), with single antennas the CZF-EGCweights are real. Thus, we just perform a preequalization am-plitude operation at the transmission side while the phaseequalization is done at the receiver. In the case of CZF, weperform the amplitude and phase equalization at transmis-sion side. For two antennas, the performance of the CZF-EGC is slightly better than the one of the CZF algorithm,while for a number of antenna elements greater than four,the performance of both algorithms is nearly identical.

The simulations leading to Figure 3 were run for Et/No =10 dB, and the metric used is the average BER as a functionof number of the users and considers the cases of a single-,two-, and four-antenna elements. From this figure we cansee that for a single antenna, the CZF algorithm performanceoutperforms the one obtained with conventional receiverequalizers for number of users up to 16 (half load).

Beyond this point the performance degradation rapidlyincreases with the number of users. This arises because for asingle antenna the difference between the number of degreesof freedom and constraints tends to zero as the number ofusers increases. When the number of antennas increases, theperformance of the CZF improves, because we increase thenumber of degrees of freedom keeping the number of con-straints. From Figure 3 we can see that for M = 2 and 4, andthe CZF gives better results than all conventional equalizers,even for full load system. Concerning CZF-EGC algorithm,we can see that performance is much better that all conven-tional receiver equalizers even for single antenna. When weincrease the number of antenna elements, the performanceof the CZF-EGC tends to the CZF performance, even for full-load system.

K = 32

M = 1

M = 1

M = 2M = 8

M = 4

CZFCZF-EGCEGC

MRCMMSEMRC single user

0 2 4 6 8 10 12 14 16 18 20 22

Et/No (dB)

10−4

10−3

10−2

10−1

Ave

rage

BE

RFigure 4: Performance comparison between the CZF, CZF-EGCand conventional receivers as function of Et/No, for beamformingcase.

The simulation results for the beamforming case areshown in Figure 4, where the simulation metrics and param-eters other than the channels correlation are identical to theones considered for Figure 2. The results show that the CZFalgorithm outperforms all the conventional equalizers, ex-cept for the single-antenna case (with loads close to full) ashappened with the diversity scenario. From this figure we cansee that the performance of the CZF with two antennas isworse than the conventional MMSE combiner. With eight-antenna elements, the CZF performance is very close to theperformance of the MRC single user. The performance of theCZF-EGC with single antenna is very good as compared withall conventional equalizers and CZF. For the single-antennacase we have the same number of degrees of freedom for bothalgorithms, but for CZF-EGC case we just perform the am-plitude equalization at transmitter side, while for CZF weperform amplitude and phase equalization. We can even seethat the performance is similar to the one obtained with CZFalgorithm for four antennas. The performance of the CZF-EGC for four and eight antennas is very close to the one ob-tained with CZF.

5. CONCLUSIONS

We proposed space-frequency preequalization techniques fordownlink TDD MC-CDMA, using antenna arrays at BS, fortwo different receivers: the conventional EGC and a simpledespread receiver without channel equalization. We analyti-cally derived the proposed preequalization algorithms, basedon a constrained zero-forcing criterion. The performancewas assessed for either of the diversity and beamforming sce-narios and compared against one of conventional receivers.

Downlink Preequalization Techniques for TDD MC-CDMA Systems 73

The results have shown that a considerable MAI reduction isobtained with the CZF-EGC technique, and with CZF whenan antenna array is used at BS. For a single antenna, the per-formance of the CZF-EGC outperforms the CZF for a full-load case, while with multiple antennas the performancesare very similar. Both techniques allow a significant improve-ment of the user capacity and move the most-demanded pro-cessing task from BS to MT, keeping this one as simple aspossible.

ACKNOWLEDGMENTS

The work presented in this paper was supported by theEuropean project IST-2001-32620, MATRICE project, andPortuguese Foundation for Science and Technology (FCT)through project POSI/CPS/46701/2002 and a grant to thefirst author.The work presented in this paper was publishedin part at VTC FALL 2003 and MC-SS 2003 proceedings.

REFERENCES

[1] S. Abeta, H. Atarashi, and M. Sawahashi, “Forward link ca-pacity of coherent DS-CDMA and MC-CDMA broadbandpacket wireless access in a multi-cell environment,” in Proc.IEEE Vehicular Technology Conference (VTC ’00), vol. 5, pp.2213–2218, Boston, Mass, USA, September 2000.

[2] IST MATRICE project, web site: http://www.ist-matrice.org.[3] S. Hara and R. Prasad, “Overview of multicarrier CDMA,”

IEEE Communications Magazine, vol. 35, no. 12, pp. 126–133,1997.

[4] H. Atarashi, N. Maeda, S. Abeta, and M. Sawahashi, “Broad-band packet wireless access based on VSF-OFCDM andMC/DS-CDMA,” in Proc. The 13th IEEE International Sym-posium on Personal, Indoor and Mobile Radio Communications(PIMRC ’02), vol. 3, pp. 992–997, Lisboa, Portugal, September2002.

[5] B. R. Vojcic and W. M. Jang, “Transmitter precoding in syn-chronous multiuser communications,” IEEE Trans. Commu-nications, vol. 46, no. 10, pp. 1346–1355, 1998.

[6] A. Silva and A. Gameiro, “Pre-filtering techniques using an-tenna arrays for downlink TDD MC-CDMA systems,” in Proc.4th International Workshop on Multi-Carrier Spread-Spectrum(MC-SS ’03), Oberpfaffenhofen, Germany, September 2003.

[7] A. Silva and A. Gameiro, “Transmit space-frequency prefilter-ing technique for downlink TDD MC-CDMA systems,” inProc. IEEE Semiannual Vehicular Technology Conference, Or-lando, Fla, USA, October 2003.

[8] A. F. Naguib, A. J. Paulraj, and T. Kailath, “Capacity improve-ment with base-station antenna arrays in cellular CDMA,”IEEE Trans. Vehicular Technology, vol. 43, no. 3, pp. 691–698,1994.

[9] C. K. Kim and Y. S. Cho, “Performance of a wireless MC-CDMA system with an antenna array in a fading channel: re-verse link,” IEEE Trans. Communications, vol. 48, no. 8, pp.1257–1261, 2000.

[10] S. Haykin, Adaptative Filter Theory, Prentice-Hall, Upper Sad-dle River, NJ, USA, 3rd edition, 1996.

[11] J. Medbo and P. Schramm, “Channel models for Hiper-lan/2 in different indoor scenarios,” ETSI/BRAN documentNo. 3ERI085b, March 1998.

Adao Silva received his B.S. and M.S. de-grees from the University of Aveiro, bothin electronics and telecommunications, in1999 and 2002, respectively. He is cur-rently working towards the Ph.D. degree atthe same university. He became an InvitedAssistant Professor in the Department ofElectronics and Telecommunications of theUniversity of Aveiro, and a Researcher atthe Instituto de Telecomunicacoes, Polo deAveiro. His main interests lie in signal processing techniques andspace-time coding for wireless communications. His current re-search activities involve preequalization and space-time-frequencyalgorithms for the broadband component of 4G systems.

Atılio Gameiro received his Licenciatura(five-year course) and his Ph.D. from theUniversity of Aveiro in 1985 and 1993,respectively. He is currently a Professorin the Department of Electronics andTelecommunications of the University ofAveiro, and a researcher at the Instituto deTelecomunicacoes, Polo de Aveiro, where heis a head of group. His main interests liein signal processing techniques for digitalcommunications and communication protocols. Within this re-search line, he has done work for optical and mobile commu-nications, at either the theoretical or experimental level, and haspublished over 100 technical papers in international journals andconferences. His current research activities involve space-time-frequency algorithms for the broadband component of 4G systemsand joint design of layers 1 and 2.

Photograph © Turisme de Barcelona / J. Trullàs

Preliminary call for papers

The 2011 European Signal Processing Conference (EUSIPCO 2011) is thenineteenth in a series of conferences promoted by the European Association forSignal Processing (EURASIP, www.eurasip.org). This year edition will take placein Barcelona, capital city of Catalonia (Spain), and will be jointly organized by theCentre Tecnològic de Telecomunicacions de Catalunya (CTTC) and theUniversitat Politècnica de Catalunya (UPC).EUSIPCO 2011 will focus on key aspects of signal processing theory and

li ti li t d b l A t f b i i ill b b d lit

Organizing Committee

Honorary ChairMiguel A. Lagunas (CTTC)

General ChairAna I. Pérez Neira (UPC)

General Vice ChairCarles Antón Haro (CTTC)

Technical Program ChairXavier Mestre (CTTC)

Technical Program Co Chairsapplications as listed below. Acceptance of submissions will be based on quality,relevance and originality. Accepted papers will be published in the EUSIPCOproceedings and presented during the conference. Paper submissions, proposalsfor tutorials and proposals for special sessions are invited in, but not limited to,the following areas of interest.

Areas of Interest

• Audio and electro acoustics.• Design, implementation, and applications of signal processing systems.

l d l d d

Technical Program Co ChairsJavier Hernando (UPC)Montserrat Pardàs (UPC)

Plenary TalksFerran Marqués (UPC)Yonina Eldar (Technion)

Special SessionsIgnacio Santamaría (Unversidadde Cantabria)Mats Bengtsson (KTH)

FinancesMontserrat Nájar (UPC)• Multimedia signal processing and coding.

• Image and multidimensional signal processing.• Signal detection and estimation.• Sensor array and multi channel signal processing.• Sensor fusion in networked systems.• Signal processing for communications.• Medical imaging and image analysis.• Non stationary, non linear and non Gaussian signal processing.

Submissions

Montserrat Nájar (UPC)

TutorialsDaniel P. Palomar(Hong Kong UST)Beatrice Pesquet Popescu (ENST)

PublicityStephan Pfletschinger (CTTC)Mònica Navarro (CTTC)

PublicationsAntonio Pascual (UPC)Carles Fernández (CTTC)

I d i l Li i & E hibiSubmissions

Procedures to submit a paper and proposals for special sessions and tutorials willbe detailed at www.eusipco2011.org. Submitted papers must be camera ready, nomore than 5 pages long, and conforming to the standard specified on theEUSIPCO 2011 web site. First authors who are registered students can participatein the best student paper competition.

Important Deadlines:

P l f i l i 15 D 2010

Industrial Liaison & ExhibitsAngeliki Alexiou(University of Piraeus)Albert Sitjà (CTTC)

International LiaisonJu Liu (Shandong University China)Jinhong Yuan (UNSW Australia)Tamas Sziranyi (SZTAKI Hungary)Rich Stern (CMU USA)Ricardo L. de Queiroz (UNB Brazil)

Webpage: www.eusipco2011.org

Proposals for special sessions 15 Dec 2010Proposals for tutorials 18 Feb 2011Electronic submission of full papers 21 Feb 2011Notification of acceptance 23 May 2011Submission of camera ready papers 6 Jun 2011