Mimo Wireless Networks. Channels, Techniques and Standards for Multi-Antenna, Multi-User and Multi-Cell Systems

752 Pages·2013·19.51 MB·English

by  Bruno Clerckx and Claude Oestges (Auth.)

MIMO Wireless
Networks
To my wife, family and friends
Bruno Clerckx
MIMO Wireless
Networks
Channels, Techniques and
Standards for Multi-Antenna,
Multi-User and Multi-Cell
Systems
Second edition
Bruno Clerckx and Claude Oestges
AMSTERDAM • BOSTON • HEIDELBERG • LONDON
NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO
Academic Press is an Imprint of Elsevier
Academic Press is an imprint of Elsevier
The Boulevard, Langford Lane, Kidlington, Oxford, OX5 1GB
225 Wyman Street, Waltham, MA 02451, USA
Second edition 2013
Copyright © 2013 Elsevier Ltd. All Rights Reserved
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Details on how to seek permission, further information about the
Publisher’s permissions policies and our arrangement with organizations such as the Copyright Clearance
Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions
This book and the individual contributions contained in it are protected under copyright by the Publisher
(other than as may be noted herein).
First Edition 2007
© 2007 Claude Oestges and Bruno Clerckx. Published by Elsevier. All rights reserved
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical treatment
may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and
using any information, methods, compounds, or experiments described herein. In using such informa-
tion or methods they should be mindful of their own safety and the safety of others, including parties for
whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any
liability for any injury and/or damage to persons or property as a matter of products liability, negligence
or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in
the material herein.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Number Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
ISBN: 978-0-12-385055-3
For information on all Academic Press publications
visit our website at store.elsevier.com
Printed and bound in the United Kingdom
13 14 15 16 10 9 8 7 6 5 4 3 2 1
List of Figures
Figure 1.1 TypicalreceivedsignalstrengthinaRayleighfadingchannel. 5
Figure 1.2 DiversitygaininRayleighfadingchannels. 8
Figure 1.3 DiversityandarraygainsinRayleighfadingchannels. 9
Figure 1.4 PerformanceoftransmitMRCandAlamoutischemeswithtwotransmit
antennasini.i.d.Rayleighfadingchannels(forBPSKmodulation). 19
Figure 1.5 PerformanceofdominanteigenmodeandAlamoutitransmissionsina
2×2i.i.d.Rayleighfadingchannel(withQPSKmodulation). 24
Figure 2.1 Atypicalmultipathscenario. 31
Figure 2.2 ExtensionofBello’sfunctionstothespatialandangulardomains. 34
Figure 2.3 TypicalDopplerspectraformobileandfixedscenarios. 39
Figure 2.4 Transmitcorrelationasafunctionofrelativeantennaspacingd /λand
t
azimuthspread(thelargerκ,thelowerthespread). 48
Figure 2.5 Mutual dipole impedance (real and imaginary parts) as a function of
antennaspacingrelativetothewavelength. 52
Figure 2.6 Field-radiation pattern (in magnitude) of the right-hand side antenna
forseveralinter-elementspacings. 53
Figure 2.7 Copolar (solid) and cross-polar (dahsed) field-radiation patterns (in
magnitude)ofaplanemonopolarantennaat2.4GHz. 55
Figure 3.1 EffectivediversitymeasureinseveralRiceanMIMOchannels. 66
Figure 3.2 PathconfigurationforscenarioA. 67
Figure 3.3 PathconfigurationforscenarioB. 67
Figure 3.4 Correlationmatrixdistancefordifferentscenarios. 68
Figure 3.5 XPDmodeling:separationbetweenspaceandpolarization. 80
Figure 3.6 Effectivediversitymeasureina2×2Kronecker-structuredsystemas
afunctionoftransmitandreceivecorrelations. 86
Figure 3.7 Virtualchannelrepresentationofa2×4channel. 87
Figure 3.8 ImagingofthescatteringenvironmentforscenarioAvia(cid:2)v,A(cid:2)(cid:2)v,A
(indBscale). 95
Figure 3.9 ImagingofthescatteringenvironmentforscenarioBvia(cid:2)v,B(cid:2)(cid:2)v,B
(indBscale). 95
xviii List of Figures
Figure 3.10 Mutualinformationvs.SNRfordifferentmodelsintwo8×8scattering
scenarios. 98
Figure 4.1 One-ringmodel. 104
Figure 4.2 Two-ringmodel(symmetric). 106
Figure 4.3 Combinedelliptical-ringmodelforthreeellipsesandalocalscatterer
ring(thesizeofthediscandthecirclehavebeenincreasedforbetter
legibility). 108
Figure 4.4 Channelcorrelations(includingtheKroneckerapproximationofs and
1
s asequaltort)vs.receiveantennaspacingd/λ. 112
2
Figure 4.5 Illustration of exponential decay of mean cluster amplitude and ray
amplitudewithinclusters. 114
Figure 4.6 GeneralstructureofCOST273model,withlocal,single-bounceand
twinclusters. 119
Figure 4.7 WINNERmodelphilosophy. 120
Figure 4.8 WINNERmodel:segmentsimulation. 121
Figure 4.9 Extendedclustermodelingthedensemultipaths. 122
Figure 4.10 ExampleofacommonclusterinascenariowithoneMTandtwoBSs
+
[PTH ss]. 123
Figure 4.11 WINNERIImulti-linkmodel. 123
Figure 5.1 Principleofwater-fillingalgorithm. 129
Figure 5.2 Capacityofvariousi.i.d.channelsat20dBSNR. 139
Figure 5.3 Ergodiccapacityofvariousn ×n i.i.d.Rayleighchannelswithfull
r t
(CSIT)andpartial(CDIT)channelknowledgeatthetransmitter. 141
Figure 5.4 Average mutual information of various correlated channels at 20 dB
SNRasafunctionofone/bothcross-channelcorrelation(s). 147
Figure 5.5 Averagemutualinformationofdiagonalchannelsat15dBSNRasa
functionofthenumberofantennasateachside. 148
Figure 5.6 Averagemutualinformationoftwofinitescattererchannelsat20dB
SNRasafunctionofthenumberofantennasateachside. 150
Figure 5.7 Mutualinformationofvariousstrategiesat0dBSNRasafunctionof
thetransmitcorrelation|t|. 152
Figure 5.8 Optimalfractionofpowertononbeamformingmodeasafunctionof
|t|andρ forn =2. 153
r
Figure 5.9 Mutual information of Ricean 2×2 channels for different K-factors
(H˜ =Hw,so K =0correspondstoaRayleighi.i.d.channel). 155
Figure 5.10 MutualinformationofRicean2×2uni-anddual-polarizedchannels. 156
Figure 5.11 Asymptoticdiversity-multiplexingtrade-offg(cid:5)(g ,∞)ini.i.d.Rayleigh
d s
fadingchannels. 161
List of Figures xix
Figure 5.12 OutageprobabilityP (R)asafunctionofthetransmissionrateRfor
out
bothfixedandvariableratescalingas R =g log (ρ)in2×2MIMO
s 2
i.i.d.Rayleighfadingchannels(courtesyofH.Yao[YW03]). 162
Figure 5.13 Diversity-multiplexing trade-off g(cid:5)(g ,ρ) at realistic SNR (5 and 10
d s
dB) of a 2×2 MIMO i.i.d. Rayleigh fading channel (courtesy of R.
Narasimhan[Nar05]). 167
Figure 5.14 Normalizedmaximumdiversitygaingˆ(cid:5)(0,ρ)/(n n )asafunctionof
d t r
SNRini.i.d.Rayleighfadingchannels. 168
Figure 5.15 Diversity-multiplexing trade-off g(cid:5)(g ,ρ) at realistic SNR (5 and 10
d s
dB)in2×2MIMOtransmitcorrelatedRayleighfadingchannels(cour-
tesyofR.Narasimhan[Nar06a]). 169
Figure 5.16 Diversity-multiplexingtrade-offg(cid:5)(g ,ρ)atrealisticSNR(10dB)in
d s
2×2MIMOtransmitcorrelatedRayleighandRicean(K =5and10
dB)fadingchannels(courtesyofR.Narasimhan[Nar06a]). 170
Figure 6.1 Generaloverviewofaspace-timeencoderofaMIMOsystem. 174
Figure 6.2 Bit error rate (BER) of Spatial Multiplexing with various receivers
(ML, (ordered) ZF SIC, ZF) in i.i.d. slow Rayleigh fading channels
withn =2andn =2for4bits/s/Hz. 205
t r
Figure 6.3 Block error rate of the Alamouti code in i.i.d. slow Rayleigh fading
channelswithn =2inR =2log (M)=4,8,12,16bits/s/Hztrans-
r 2
missions using M2-QAM constellations of sizes M = 4,16,64,256
(courtesyofH.Yao[YW03]). 213
Figure 6.4 Bit error rate (BER) of several LDC in i.i.d. slow Rayleigh fading
channelswithn =2andn =2for4bits/s/Hz. 217
t r
Figure 6.5 Block error rate of the tilted-QAM code in i.i.d. slow Rayleigh fad-
ing channels with n = 2 in a R = 2n log (M) = 4,8,...,32-
r t 2
bit/s/Hz transmissions using M2-QAM constellations of sizes M =
2,4,8,...,256(courtesyofH.Yao[YW03]). 220
Figure 6.6 Biterrorrate(BER)ofDayalandAlamouticodesini.i.d.slowRayleigh
fadingchannelswithn =1andn =2ina4-bit/s/Hztransmission. 221
r r
Figure 6.7 Asymptotic diversity-multiplexing trade-off g (g ,∞) achieved by
d s
severalspace-timecodesina2×2i.i.d.RayleighfadingMIMOchannel. 223
Figure 6.8 Biterrorrate(BER)ofseveralalgebraicspace-timeblockcodesini.i.d.
slowRayleighfadingchannelswithn =2andn =2ina4-bit/s/Hz
t r
transmission. 224
Figure 6.9 Bit error rate (BER) of SM, Dayal and Alamouti codes in i.i.d. slow
Rayleighfadingchannelswithn =2,3,4ina4-bit/s/Hztransmission. 225
r
Figure 6.10 LabelingoftheQPSKconstellation. 226
Figure 6.11 STTCencoderfortwotransmitantennas. 227
xx List of Figures
Figure 6.12 Trellis representation of QPSK 4-state 2 bits/s/Hz space-time trellis
codesfortwotransmitantennas:(a)“TSC”code(delay-diversitycode)
[Wit93,SW94,TSC98]; (b) “BBH” code [BBH00]; (c) “CYV” code
[CYV01];(d)“FVY”code[FVY01]. 228
Figure 6.13 TrellisrepresentationofaQPSK8-state2-transmitantennaspace-time
trelliscodewhosegeneratormatrixisgivenby(6.201). 229
Figure 6.14 Frameerrorrateofseveral4-stateSTTCini.i.d.slowRayleighfading
channelswithn =2andn =2,4. 233
t r
Figure 6.15 Frameerrorrateof4-stateand8-state“CYV”and“TSC”codesini.i.d.
slowRayleighfadingchannelswithn =2andn =4. 234
t r
Figure 6.16 Frameerrorrateofseveral4-stateSTTCini.i.d.fastfadingchannels
withn =2andn =1. 235
t r
Figure 7.1 Biterrorrate(BER)ofSpatialMultiplexingwiththreereceiversini.i.d.
Rayleighslowfadingchannelswithn =n =4andQPSK. 242
t r
Figure 7.2 (a)Thesphereiscenteredatthereceivedvectorandcontainsthelattice
pointstobeenumerated.(b)Thesphereistransformedintoanellipsoid
intheT coordinatesystem. 245
Figure 7.3 Illustrationofspheredecodingtree. 247
Figure 7.4 Illustrationofbreadth-firstsearchtreefor K =1. 251
Figure 7.5 Principleofslowestdescentdetection:theellipsoidsrepresenttheequi-
probable regions of logp(y|s), and e is chosen as the direction of
SD
slowestdecent. 254
Figure 7.6 BERperformanceofslowestdescentdetectionini.i.d.Rayleighslow
fadingchannelswithn =n =4andQPSK. 256
t r
Figure 7.7 BICMiterativereceiver. 258
Figure 7.8 Kalmanfilter-basedestimationcombinedwithMMSE-DFEreceiver. 261
Figure 8.1 Visualizationoftheimpactofthescatteringrichnessandinter-element
spacingonMIMOsystemperformance. 266
Figure 8.2 G (θ |c ) (θ varying over 360◦) for the four possible phase shifts
t t k t
betweentwotransmittedQPSKsymbolsandMTinterelementspacing
d /λ=0.1. 268
t
Figure 8.3 G (θ |c ) (θ varying over 360◦) for the four possible phase shifts
t t k t
betweentwotransmittedQPSKsymbolsandMTinterelementspacing
d /λ=0.5. 269
t
Figure 8.4 Symbolerrorrateasafunctionofthephaseshift(inradians)between
thetransmittedQPSKsymbols. 270
Figure 8.5 Performance of full rank LDCs in i.i.d. and correlated channels with
n =2andn =2. 280
t r
Figure 8.6 Performance of rank-deficient LDCs in i.i.d. and spatially correlated
channelswithn =2andn =2. 288
t r
List of Figures xxi
Figure 8.7 PerformanceofSTTCsini.i.d.andspatiallycorrelatedchannelswith
n =2andn =4. 290
t r
Figure 8.8 Performance ofSTTCsini.i.d.andspatiallycorrelatedfastRayleigh
fadingchannelswithn =2andn =1. 291
t r
Figure 8.9 Performanceofrank-deficientSpatialMultiplexingschemeson2×2
Riceanfadingchannelswith K =4and K =10. 292
Figure 8.10 Performanceofapproximatelyuniversalcodeson2×2Riceanfading
channelswith K =4and K =10. 293
Figure 9.1 Repetitioncodingfor L =2withrate R =4bits/s/Hz. 302
Figure 9.2 Permutationcodingfor L =2withrate R =4bits/s/Hz. 302
Figure 9.3 MutualcouplingeffectsontheG (θ |C,a (θ ))ofaSMschemewith
sum t t t
QPSK. 312
Figure 9.4 MutualcouplingeffectsontheSERofaSMschemewithQPSK. 313
Figure 9.5 G (θ |C,a (θ ))ofseveralSMschemesasafunctionoftheangleof
sum t t t
departureθ [rad]. 321
t
Figure 9.6 SERasafunctionofthephaseoft for|t| = 0.95(up)andSERasa
functionof|t|withthephaseoftequalto0(down). 322
Figure 9.7 Biterrorrateofseveral2×2SMschemesini.i.d.andspatiallycorre-
latedRayleighfadingchannels(withtwoantennaorientationsθ = 0
t
andθ =0.63). 323
t
Figure 9.8 G (θ |C,a (θ ))ofseveralLDCasafunctionoftheangleofdepar-
sum t t t
tureθ [rad]. 324
t
Figure 9.9 G (θ |C,a (θ ))ofseveralLDCsasafunctionoftheangleofdepar-
sum t t t
tureθ [rad]. 325
t
Figure 9.10 G (θ |C,a (θ ))ofseveralfull-ratealgebraiccodesasafunctionof
sum t t t
theangleofdepartureθ [rad]. 326
t
Figure 9.11 G (θ |C,a (θ ))ofseveral4-and8-stateSTTCsasafunctionofthe
sum t t t
angleofdepartureθ . 327
t
Figure 9.12 Frameerrorrateofseveral4-stateSTTCsini.i.d.andspatiallycorre-
latedRayleighfadingchannelswithn =2andn =4. 328
t r
Figure 9.13 Frameerrorrateofseveral8-stateSTTCsini.i.d.andspatiallycorre-
latedRayleighfadingchannelswithn =2andn =4. 329
t r
Figure 9.14 G (θ |C,a (θ ))ofseveral4-stateSTTCsasafunctionoftheangle
sum t t t
ofdepartureθ . 331
t
Figure 9.15 G (θ |C,a (θ )) of several 4-states STTCs as a function of the
product t t t
angleofdepartureθ . 332
t
Figure 9.16 Frameerrorrateofseveral4-statesSTTCsini.i.d.andspatiallycorre-
latedRayleighfastfadingchannelswithn =2andn =1. 333
t r
xxii List of Figures
(cid:5)(cid:5)
Figure 10.1 OverviewofprecoderP:Wactsasamulti-modebeamformer,C is
thecodewordbeingshapedbyS1/2. 337
Figure 10.2 PerformanceofatransmitcorrelationbasedprecodedAlamoutischeme
in2×2transmitcorrelated(t =0.7)Rayleighchannels. 343
Figure 10.3 PerformanceofatransmitcorrelationbasedprecodedAlamoutischeme
in2×2transmitcorrelated(t =0.95)Rayleighchannels. 344
Figure 10.4 Performanceofa4-stateSTTC“CYV”inacorrelatedRayleighfading
channel (n = n = 2, n = 4), with a high transmit correlation
e t r
usingtwoprecodingschemes:UE˜· =Int andE˜· =argminE˜(cid:6)=0det(E˜)
(proposedin[SP02]). 350
Figure 10.5 Bit error rate of Spatial Multiplexing as a function of the transmit
correlationcoefficientt in2×2correlatedMIMOchannelswithand
withoutprecoding(SNR=15dB). 356
Figure 10.6 BiterrorrateofSpatialMultiplexingincorrelatedchannelswithand
without precoding: precoders I, II and III exploit the knowledge of
t, while the robust precoder has been designed following the G
sum
criterioninChapter9. 357
Figure 10.7 IllustrationofGrassmannian,adaptiveanddifferentialcodebooks. 362
Figure 10.8 Symbolerrorrate(SER)ofa3×3MIMOsystemusing2-bitand6-bit
quantized BPSK-based dominant eigenmode transmissions (courtesy
ofD.Love[LHS03]). 365
Figure 10.9 Symbol error rate (SER) of a 8×1 MISO system using 6-bit quan-
tizedi.i.d.androtateddominanteigenmodetransmissionsinspatially
correlatedRayleighchannels(courtesyofD.Love[LH06]). 368
Figure 10.10 Normalized average distortion (SNR loss) df,n as a function of the
codebooksizen =2B andthetransmitcorrelationcoefficientt with
p
n =4. 370
t
Figure 10.11 Symbolerrorrate(SER)of3-bitand6-bitprecodedAlamoutischemes
in2×4i.i.d.Rayleighfadingchannels(courtesyofD.Love[LH05a]). 378
Figure 10.12 Symbolvectorerrorrate(SVER)ofa6-bitprecodedBPSK-basedSM
scheme in i.i.d. Rayleigh fading channels with n = 4 and n = 2
t r
(courtesyofD.Love[LH05b]). 381
Figure 11.1 OFDMmodulatoranddemodulator. 386
Figure 11.2 BlockdiagramofaMIMO-OFDMsystem. 389
Figure 11.3 FER of the 16-state “FVY” code for L = 2, 3 and 4 in uniformly
distributedi.i.d.Rayleighchannelswithandwithoutinterleaver. 406
Figure 11.4 CyclicdelaydiversityinMIMO-OFDM. 409
Figure 12.1 MIMOBroadcastChannel(BC)andMultipleAccessChannel(MAC). 420