`i a pe Rohit Leaeeee Gore
`
`IPR2018-01477
`Apple Inc. EX1009 Page 1
`
`IPR2018-01477
`Apple Inc. EX1009 Page 1
`
`
`
`Introduction to Space-Time
`Wireless Communications
`
`Arogyaswami Paulraj
`
`Stanford University
`
`Rohit Nabar
`
`ETH, Zurich
`
`Dhananjay Gore
`
`Stanford University
`
`5 CAMBRIDGE
`[)4\'Gs)
`i UNIVERSITY PRESS
`
`
`
`IPR2018-01477
`Apple Inc. EX1009 Page 2
`
`IPR2018-01477
`Apple Inc. EX1009 Page 2
`
`
`
`
`
`PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
`
`ThePitt Building, Trumpington Street, Cambridge, United Kingdom
`CAMBRIDGE UNIVERSITY PRESS
`
`The Edinburgh Building, Cambridge CB2 2RU, UK
`40 West 20th Street, New York, NY 10011-4211, USA
`477 Williamstown Road, Port Melbourne, VIC 3207, Australia
`Ruiz de Alarcon 13, 28014 Madrid, Spain
`Dock House, The Waterfront, Cape Town 8001, South Africa
`
`http://www.cambridge.org
`
`© Cambridge University Press 2003
`
`This bookis in copyright. Subjectto statutory exception
`and to the provisions of relevantcollective licensing agreements,
`no reproduction of any part may take place without
`the written permission of Cambridge University Press.
`
`First published 2003
`
`Printed in the United Kingdom at the University Press, Cambridge
`
`Typefaces Times 10.5/14 pt and Helvetica Neve
`
`System ISTEX2¢
`
`[TB]
`
`A catalog recordfor this bookis available fromthe British Library
`
`ISBN 0 521 82615 2 hardback
`
`IPR2018-01477
`Apple Inc. EX1009 Page 3
`
`IPR2018-01477
`Apple Inc. EX1009 Page 3
`
`
`
`Contents
`
`List offigures
`List oftables
`Preface
`List of abbreviations
`List of symbols
`
`1
`
`Introduction
`
`1.1
`
`1.2
`
`1.3
`
`History of radio, antennas and array signal processing
`Exploiting multiple antennas in wireless
`1.2.1 Array gain
`1.2.2 Diversity gain
`1.2.3 Spatial multiplexing (SM)
`1.2.4 Interference reduction
`ST wireless communication systems
`
`2
`
`ST propagation
`
`2.1
`
`Introduction
`
`22
`
`The wireless channel
`
`2.2.1 Path loss
`2.2.2 Fading
`Scattering model in macrocells
`Channel as a ST random field
`2.4.1 Wide sense stationarity (WSS)
`2.4.2 Uncorrelated scattering (US)
`2.4.3 Homogeneous channels (HO)
`Scattering functions
`
`2.3
`
`2.4
`
`io)
`
`vii
`
`page xiv
`XXil
`
`XXILl
`
`XXVI
`
`XXX
`
`
`
`woooconnnr
`
`ll
`
`il
`
`11
`
`12
`
`18
`
`20
`
`22
`
`22
`
`23
`
`24
`
`IPR2018-01477
`Apple Inc. EX1009 Page 4
`
`IPR2018-01477
`Apple Inc. EX1009 Page 4
`
`
`
`
`
`Contents_ viil
`
`—l
`
`WwNmWwbdOoco
`‘nynwtadWwWwWMmhwhoho
`
`tf©)Lo
`
`2.6
`
`2.7
`
`2.8
`
`2.9
`
`Polarization and field diverse channels
`Antennaarray topology
`Degenerate channels
`Reciprocity and its implications
`
`3
`
`ST channel and signal models
`
`3:
`
`—_
`
`Introduction
`
`O12
`
`Definitions
`
`3.2.1 SISO channel
`
`3.2.2 SIMO channel
`
`3.2.3 MISO channel
`
`3.3
`
`3.2.4 MIMO channel
`Physical scattering model for ST channels
`3.3.1 SIMO channel
`
`3.4
`
`3. Cn
`
`3.6
`
`Sp
`
`3.8
`
`3.9
`
`3.3.2 MISO channel
`
`3.3.3 MIMO channel
`
`Extended channel models
`3.4.1 Spatial fading correlation
`3.4.2 LOS component
`3.4.3 Cross-polarized antennas
`3.4.4 Degenerate channels
`Statistical properties of H
`3.5.1 Singular values of H
`3.5.2 Squared Frobenius norm of H
`Channel measurements andtest channels
`
`Sampled signal model
`3.7.1 Normalization
`3.7.2 SISO sampled signal model
`3.7.3 SIMO sampledsignal model
`3.7.4 MISO sampled signal model
`3.7.5 MIMO sampledsignal model
`ST multiuser and ST interference channels
`
`3.8.1 ST multiuser channel
`3.8.2 ST interference channel
`ST channel estimation
`3.9.1 Estimating the ST channelat the receiver
`3.9.2 Estimating the ST channelat the transmitter
`
`ER
`
`FORLANI
`
`IPR2018-01477
`Apple Inc. EX1009 Page 5
`
`IPR2018-01477
`Apple Inc. EX1009 Page 5
`
`
`
`x
`Bee
`
`Contents
`
`EEESl
`
`4
`
`Capacity of ST channels
`
`4.1 Introduction
`4.2 Capacity of the frequency flat deterministic MIMO channel
`4.3. Channel unknown to the transmitter
`
`4.4 Channel known to the transmitter
`4.4.1 Capacities of SIMO and MISO channels
`4.5 Capacity of random MIMOchannels
`4.5.1 Capacity of H,, channels for large M
`4.5.2 Statistical characterization of the information rate
`4.6 Influence of Ricean fading, fading correlation, XPD and degeneracy on
`MIMO capacity
`4.6.1 Influence of the spatial fading correlation
`4.6.2 Influence of the LOS component
`4.6.3 Influence of XPD in a non-fading channel
`4.6.4 Influence of degeneracy
`4.7 Capacity of frequency selective MIMO channels
`
`ee)
`
`5
`
`Spatial diversity
`
`5.1 Introduction
`5.2 Diversity gain
`5.2.1 Coding gain vs diversity gain
`5.2.2 Spatial diversity vs time/frequency diversity
`5.3 Receive antennadiversity
`5.4 Transmit antennadiversity
`5.4.1 Channel unknownto the transmitter: MISO
`5.4.2 Channel known to the transmitter: MISO
`5.4.3 Channel unknown to the transmitter: MIMO
`5.4.4 Channel knownto the transmitter: MIMO
`5.5 Diversity order and channel variability
`5.6 Diversity performance in extended channels
`5.6.1 Influence ofsignal correlation and gain imbalance
`5.6.2 Influence of Ricean fading
`5.6.3 Degenerate MIMO channels
`5.7 Combined space andpath diversity
`
`63
`
`63
`63
`65
`
`66
`70
`71
`7\
`72
`
`en
`77
`78
`80
`80
`81
`
`86
`
`86
`86
`89
`90
`90
`92
`93
`95
`o7
`98
`100
`102
`102
`104
`105
`106
`
`IPR2018-01477
`Apple Inc. EX1009 Page 6
`
`IPR2018-01477
`Apple Inc. EX1009 Page 6
`
`
`
`Contents
`X
`et
`
`5.8 Indirect transmit diversity
`5.8.1 Delay diversity
`5.8.2. Phase-roll diversity
`5.9 Diversity of a space-time-frequency selective fading channel
`
`6
`
`ST coding without channel knowledgeat transmitter
`
`6.1 Introduction
`6.2 Coding andinterleaving architecture
`6.3 ST coding for frequency flat channels
`6.3.1 Signal model
`6.3.2 ST codeword design criteria
`6.3.3 ST diversity coding (r, < 1)
`6.3.4 Performance issues
`6.3.5 Spatial multiplexing as a ST code (7, = M7)
`6.3.6 ST coding for intermediate rates (1 < r, < Mr)
`6.4 ST coding for frequency selective channels
`6.4.1 Signal model
`6.4.2 ST codeword designcriteria
`
`
`
`7 ST receivers
`
`7.1 Introduction
`
`7.2 Receivers: SISO
`7.2.1 Frequency flat channel
`7.2.2 Frequencyselective channel
`7.3 Receivers: SIMO
`7.3.1 Frequency flat channel
`7.3.2 Frequency selective channels
`7.4 Receivers: MIMO
`7.4.1 ST diversity schemes
`7.4.2 SM schemes
`7.4.3 SM with horizontal and diagonal encoding
`7.4.4 Frequency selective channel
`7.5 Iterative MIMO receivers
`
`.
`
`108
`
`108
`
`108
`
`109
`
`112
`
`Lis
`
`114
`
`114
`
`115
`
`117
`
`123
`
`123
`
`126
`
`129
`
`[29
`
`131
`
`137
`
`137
`
`137
`
`137
`
`138
`
`143
`
`143
`
`144
`
`148
`
`148
`
`149
`
`158
`
`159
`
`159
`
`IPR2018-01477
`Apple Inc. EX1009 Page 7
`
`IPR2018-01477
`Apple Inc. EX1009 Page 7
`
`
`
`Contents
`xi
`Ea
`
`as
`
`8
`
`Exploiting channel knowledgeat the transmitter
`
`8.1 Introduction
`8.2 Linear pre-filtering
`8.3 Optimal pre-filtering for maximumrate
`8.3.1 Full channel knowledge
`8.3.2 Partial channel knowledge
`8.4 Optimal pre-filtering for error rate minimization
`8.4.1 Full channel knowledge
`8.4.2 Partial channel knowledge
`8.5 Selection at the transmitter
`
`8.5.1 Selection between SM anddiversity coding
`8.5.2 Antenna selection
`
`8.6 Exploiting imperfect channel knowledge
`
`SSS)
`
`9
`
`ST OFDM and spread spectrum modulation
`
`9.1 Introduction
`
`9.2 SISO-OFDM modulation
`9.3 MIMO-OFDM modulation
`9.4 Signaling and receivers for MIMO-OFDM
`9.4.1 Spatial diversity coding for MIMO-OFDM
`9.4.2 SM for MIMO-OFDM
`9.4.3 Space-frequency coded MIMO-OFDM
`9.5 SISO-SS modulation
`9.5.1 Frequency flat channel
`9.5.2 Frequency selective channel
`9.6 MIMO-SS modulation
`9.7 Signaling and receivers for MIMO-SS
`9.7.1 Spatial diversity coding for MIMO-SS
`9.7.2 SM for MIMO-SS
`
`SSS
`
`10
`
`MIMO-multiuser
`
`10.1
`
`Introduction
`
`163
`
`163
`163
`165
`165
`166
`168
`168
`168
`17]
`
`171
`172
`
`175
`
`178
`
`178
`
`178
`182
`184
`184
`186
`186
`188
`188
`191
`193
`194
`194
`197
`
`199
`
`199
`
`IPR2018-01477
`Apple Inc. EX1009 Page 8
`
`IPR2018-01477
`Apple Inc. EX1009 Page 8
`
`
`
`xii
`PC
`
`Contents
`
`10.2 MIMO-MAC
`10.2.1
`Signal model
`10.2.2 Capacity region
`10.2.3. Signaling and receiver design
`10.3 MIMO-BC
`10.3.1
`Signal model
`10.3.2 Forward link capacity
`10.3.3. Signaling and receiver design
`10.4 Outage performance of MIMO-MU
`10.4.1 MU vs SU - single cell
`10.4.2 MU single cell vs SU multicell
`10.5 MIMO-MU with OFDM
`10.6
`CDMAand multiple antennas
`
`ea
`
`11
`
`ST co-channelinterference mitigation
`
`11.1
`
`Introduction
`
`11.2 CCI characteristics
`11.3 Signal models
`11.3.1
`SIMO interference model (reverse link)
`11.3.2. MIMOinterference channel (any link)
`11.3.3. MISO interference channel (forward link)
`11.4 CCI mitigation on receive for SIMO
`11.4.1
`Frequencyflat channel
`11.4.2 Frequency selective channel
`11.5 CCI mitigating receivers for MIMO
`11.5.1 Alamouti coded signal and interference (My; = 2)
`11.6 CCI mitigation on transmit for MISO
`11.6.1 Transmit-MRC or matched beamforming
`11.6.2 Transmit ZF or nulling beamformer
`11.6.3 Max SINR beamforming with coordination
`Joint encoding and decoding
`11.7
`11.8 SS modulation
`
`11.8.1
`ST-RAKE
`11.8.2 ST pre-RAKE
`11.9 OFDM modulation
`11.10 Interference diversity and multiple antennas
`
`201
`201
`202
`207
`208
`208
`208
`209
`213
`214
`215
`216
`216
`
`218
`
`218
`
`219
`219
`220
`222
`223
`224
`224
`226
`228
`229
`230
`230
`231
`232
`233
`233
`
`234
`235
`237
`237
`
`IPR2018-01477
`Apple Inc. EX1009 Page 9
`
`
`
`IPR2018-01477
`Apple Inc. EX1009 Page 9
`
`
`
`xiii
`=a
`
`Contents
`
`aay
`
`12
`
`Performancelimits and tradeoffs in MIMO channels
`
`12.1
`Introduction
`12.2 Error performance in fading channels
`12.3.
`Signaling rate vs PER vs SNR
`12.4
`Spectral efficiency of ST coding/receiver techniques
`12.4.1 D-BLAST
`
`12.4.2 OSTBC
`12.4.3
`ST receivers for SM
`12.4.4 Receiver comparison: Varying Mr/Mer
`System design
`12.5
`12.6 Comments on capacity
`
`References
`Index of commonvariables
`Subject index
`
`240
`
`240
`240
`24]
`244
`244
`
`245
`246
`249
`250
`251
`
`254
`271
`272
`
`IPR2018-01477
`Apple Inc. EX1009 Page 10
`
`IPR2018-01477
`Apple Inc. EX1009 Page 10
`
`
`
`5
`
`List of symbols
`
`
`rank of the matrix A
`
`real field
`
`real and imaginary parts of A, respectively
`trace of A
`lifx>O0,x.eER
`;
`.
`—
`unit step function, defined as u(x) = 0 ifx<0.xeR
`stacks A into a vector columnwise!
`
`r(A)
`
`R R
`
`{A}, {A}
`Tr(A)
`
`u(x)
`
`vec(A)
`
`x ifx>OQxeR
`ES14 is 20,xeR
`—
`(x)4
`integerfield
`
`defined as (x), =
`
`z
`
`
`
`‘IFA =[a) a2 ---
`
`a,] ism xn, then vec(A) = [al af --. al)’ ismn x 1.
`
`IPR2018-01477
`Apple Inc. EX1009 Page 11
`
`IPR2018-01477
`Apple Inc. EX1009 Page 11
`
`
`
`‘Introduction
`
`a1
`
`The radio age began over a 100 years ago with the invention of the radiotelegraph by
`Guglielmo Marconi and the wireless industry is now set for rapid growth as we enter a
`new century and a new millennium. Therapid progressin radio technologyis creating
`new and improvedservices at lower costs, whichresults in increases in air-time usage
`and the numberof subscribers. Wireless revenues are currently growing between 20%
`and 30% per year, and these broad trendsare likely to continue for several years.
`Multiple access wireless communicationsis being deployed for both fixed and mobile
`applications. In fixed applications, the wireless networks provide voiceordata for fixed
`subscribers. Mobile networks offering voice and data services can be divided into two
`classes: high mobility, to serve high speed vehicle-borne users, and low mobility, to
`serve pedestrian users. Wireless system designers are faced with a numberofchallenges.
`These include the limited availability of the radio frequency spectrum and a complex
`time-varying wireless environment (fading and multipath). In addition, meeting the
`increasing demandfor higherdata rates, better quality of service (QoS), fewer dropped
`calls, higher network capacity and user coverage calls for innovative techniques that
`improve spectral efficiency and link reliability. The use of multiple antennas at the
`receiver and/or transmitter in a wireless system, popularly known as space-time (ST)
`wireless or multiantenna communications or smart antennas is an emerging technology
`that promises significant improvements in these measures. This bookis an introduction
`to the theory of ST wireless communications.
`
`1.1.
`
`History of radio, antennas andarray signal processing
`
`The origins of radio date back to 1861 when Maxwell, while at King’s College in
`London, proposed a mathematical theory of electromagnetic (EM) waves. A practical
`demonstration of the existence of such waves was performed by Hertz in 1887 at the
`University of Karlsruhe, using stationary (standing) waves. Following this, improve-
`ments in the generation and reception of EM waves were pursued by many researchers
`in Europe. In 1890, Branly in Paris developed a “coherer” that could detectthe presence
`of EM wavesusingiron filings in a glass bottle. The coherer was further refined by
`
`IPR2018-01477
`Apple Inc. EX1009 Page 12
`
`IPR2018-01477
`Apple Inc. EX1009 Page 12
`
`
`
`Introduction
`1
`
`
`Righi at the University of Bologna and Lodge in England. Other contributions came
`from Popovin Russia, whois credited with devising the first radio antenna during his
`attempts to detect EM radiation fromlightning.
`In the summerof 1895, Marconi, at the age of 21, was inspired by the lectures on radio
`wavesby Righi at the University of Bologna and he built and demonstratedthefirst radio
`telegraph. He used Hertz’s spark transmitter, Lodge’s coherer and added antennasto as-
`semble his instrument.In 1898, Marconi improvedthe telegraph by adding a four-circuit
`tuning device, allowing simultaneous use of tworadio circuits. That year, his signal
`bridged the English Channel, 52 km wide, between Wimereux and Dover. His other tech-
`nical developments aroundthis time included the magnetic detector, which was an im-
`provementovertheless efficient coherer, the rotatory spark and the useof directive an-
`tennasto increasethesignal level andto reduce interference in duplex receivercircuits.
`In the next few years, Marconi integrated many new technologiesinto his increasingly
`sophisticated radio equipment, including the diode valve developed by Fleming, the
`crystal detector, continuous wave (CW)transmission developed by Poulsen, Fessenden
`and Alexanderson, and the triode valve or audio developed by Forrest.
`Civilian use of wireless technology began with the installation of the first 2 MHz
`land mobile radiotelephone system in 1921 by the Detroit Police Department forpolice
`car dispatch. The advantages of mobile communications were quickly realized, but its
`wider use was limited by the lack of channels in the low frequency band. Gradually,
`higher frequency bands were used, opening up the use of more channels. A key ad-
`vance was madein 1933, when Armstronginvented frequency modulation (FM), which
`madepossible high quality radio communications. In 1946, a Personal Correspondence
`System introduced by Bell Systems began service and operated at 150 MHz with speech
`channels 120 kHz apart. As demand for public wireless services began to grow, the
`Improved Mobile Telephone Service (IMTS) using FM technology was developed by
`AT&T. These werethe first mobile systems to connect with the public telephone net-
`work using a fixed numberof radio channels in a single geographic area. Extending
`such technologyto a large numberofusers with full duplex channels needed excessive
`bandwidth. A solution was found in the cellular concept (knownascellularization)
`conceived by Ring at Bell Laboratories in 1947. This concept required dividing the
`service area into smaller cells, and using a subsetof the total available radio channels
`in each cell. AT&T proposed the first high capacity analog cellular telephone system
`called the Advanced Mobile Phone Service (AMPS) in 1970. Mobile cellular systems
`have evolved rapidly since then, incorporating digital communication technology and
`serve nearly one billion subscribers worldwide today. While the Global System for
`Mobile (GSM)standard developed in Europe has gathered the largest market share,
`cellular networks in the USA haveusedthe IS-136 (using time division multiple access
`or TDMA)and IS-95 (using Code Division Multiple Access or CDMA)standards. With
`increasing use of wireless internetin the late 1990s, the demandfor higher spectraleffi-
`ciency anddata rates has led to the developmentofthe so called Third Generation (3G)
`
`IPR2018-01477
`Apple Inc. EX1009 Page 13
`
`1CNeer
`
`
`
`IPR2018-01477
`Apple Inc. EX1009 Page 13
`
`
`
`1.1 History of radio, antennas andarray signal processing
`
`Table 1.1. Performance goals for antennasin wireless
`communications
`
`
`
`AOAestimationAntennadesign Link performance
`
`
`
`Gain
`Bandwidth
`Radiation pattern
`Size
`
`Error variance
`Bias
`Resolution
`
`Coverage
`Quality
`Interference reduction
`Spectral efficiency
`
`Active integrated
`
`Phased arrays
`
`1960s
`
`Patch
`
`1950s
`
`1980s
`
`
`
`
`Hertz/Marconi/Popov
`1880-1890s
`
`Yagi-Uda
`
`1920s
`
`Directive
`
`1900s
`
`Figure 1.1: Developments in antenna (EM) technology.
`
`wireless technologies. 3G standardization failed to achieve a single common world-
`wide standard and now offers UMTS (wideband CDMA) and | XRTT asthe primary
`standards. Limitations in the radio frequency (RF) spectrum necessitate the use of
`innovative techniques to meet the increased demandin data rate and QoS.
`The use of multiple antennasat the transmitter and/or receiver in a wireless commu-
`nication link opens a new dimension — space, whichif leveraged correctly can improve
`performance substantially. Table 1.1 details the three main areasofstudy in thefield of
`radio antennasandtheir applications. Thefirst covers the electromagnetic design ofthe
`antennas and antenna arrays. The goals here are to meet design requirements for gain,
`polarization, beamwidth, sidelobelevel, efficiency and radiation pattern. The second
`area is the angle-of-arrival (AOA) estimation and, as the name indicates, focuses on
`estimating arrival angles of wavefronts impinging on the antenna array with minimum
`error and high resolution. The third area of technologythat this book focuses onis the
`use of antenna arrays to improve spectralefficiency, coverage and quality of wireless
`links.
`A timeline of the key developmentsin thefield of antenna designis given in Fig.1.1.
`The original antenna design work came from Marconi and Popov amongothersin the
`early 1900s. Marconi soon developed directional antennasfor his cross-Atlantic links.
`Antennadesign improved in frequency of operation and bandwidthin the early part of
`the twentieth century. An important breakthrough was the Yagi-Uda arraysthat offered
`high bandwidth and gain. Another important development wasthe patch antenna that
`offers low profile and cost. The use of antennas in arrays began in World WarII, mainly
`
`IPR2018-01477
`Apple Inc. EX1009 Page 14
`
`IPR2018-01477
`Apple Inc. EX1009 Page 14
`
`
`
`Introduction
`1
`
`
`Multiple source
`
`
`
`
`
`
`1985
`
`ESPRIT
`
`MUSIC
`
`1980
`
`
`
`
`
`
`Figure 1.2: Developments in AOA estimation.
`
`Maximum likelihood
`
`1964
`
`Single source
`
`Wullenweber
`
`1935
`
`Sweeny
`
`1970
`
`Adcock
`
`1919
`
`Loop
`
`for radar applications. Array design brought many newissuesto the fore, suchas gain,
`beamwidth, sidelobe level, and beamsteering.
`The area of AOAestimation had its beginnings in World War I whenloop antennas
`were used to estimate signal direction (see Fig. 1.2 for a timeline of AOA technol-
`ogy). Adcock antennas were a significant advance and were used in World WarII.
`Wullenweberarrays were developed in 1938 for lower frequencies and where accuracy
`was important, and are used in aircraft localization to this day. These techniques ad-
`dressed the single source signal wavefront case. If there are multiple sources in the same
`frequency channel or multipath arrivals from a single source, new techniques are needed.
`The problem of AOAestimation in the multisource case was properly addressed in the
`1970s and 1980s. Capon’s method [Caponetal., 1967], a well-known technique, offered
`reasonable resolution performance althoughit suffered from bias even in asymptoti-
`cally large data cases. The multiple signal classification (MUSIC) technique proposed
`by Schmidt in 1981 was a major breakthrough. MUSICis asymptotically unbiased and
`offers improved resolution performance. Later a method called estimation of signal
`parameters via rotational invariance techniques (ESPRIT) that has the remarkable ad-
`vantage of not needing exact characterization of the array manifold and yet achieves
`optimal performance was proposed [Paulraj et al., 1986; Roy et al., 1986].
`The third area of antenna applications in wireless communications is link enhance-
`ment (see Fig. 1.3). The use of multiple receive antennas for diversity goes back to
`Marconi and the early radio pioneers. So does the realization that steerable receive
`antenna arrays can be used to mitigate co-channel interference in radio systems. The
`use of antenna arrays was an active reseach area during and after World WarII in radar
`systems. More sophisticated applicationsof adaptive signal processing at the wireless
`receiver for improving diversity and interference reduction had to wait until the 1970s
`for the arrival ofdigital signal processors at which point these techniques were vigor-
`ously developed for military applications. The early 1990s saw new proposals for using
`antennas to increase capacity of wirelesslinks. Roy and Ottersten in 1996 proposed the
`use of base-station antennas to support multiple co-channelusers, Paulraj and Kailath in
`
`IPR2018-01477
`Apple Inc. EX1009 Page 15
`
`me
`
`IPR2018-01477
`Apple Inc. EX1009 Page 15
`
`
`
`1.1 History of radio, antennas andarray signal processing
`Se
`
` Adaptive
`Rx ST techniques
`
`1900s
`
`Tx-Rx ST techniques
`
`
`Jammer cancellation
`1980s
`
`2000
`
`
`
`
`
`
`
`
`Non-adaptive
`
`Butler
`
`Marconi
`
`1905
`
`Figure 1.3: Developmentsin antenna technologyfor link performance.
`
`w
`
`Oo on
`
`=Rh_~onnmon
`
`
`
`Datarate(Mbps)
`
`SNR (dB)
`
`Figure 1.4: Data rate (at 95% reliability) vs SNR for different antenna configurations. Channel
`bandwidth is 200 KHz.
`
`1994 proposed a techniquefor increasing the capacity of a wireless link using multiple
`antennasat both the transmitter and the receiver. These ideas along with the fundamen-
`tal research doneat Bell Labs [Telatar, 1995; Foschini, 1996; Foschini and Gans, 1998;
`Tarokh et al., 1998] began a new revolution in information and communications theory
`in the mid 1990s. The goalis to approach performancelimits and to explore efficient but
`pragmatic coding and modulation schemesfor wireless links using multiple antennas.
`Clearly much more work has yet to be done and thefield is attracting considerable
`research talent.
`The leverage of ST wireless technologyis significant. Figure 1.4 plots the maximum
`error-free data rate in a 200 KHz fading channel vs the signal to noise ratio (SNR)
`
`IPR2018-01477
`Apple Inc. EX1009 Page 16
`
`IPR2018-01477
`Apple Inc. EX1009 Page 16
`
`
`
`1
`
`Introduction
`
`6
`
`Es
`
`Figure 1.5: Antenna configurations in ST wireless systems (Tx: Transmitter, Rx: Receiver).
`
`that is guaranteed at 95% reliability. Assuming a target receive SNR of 20 dB, current
`single antenna transmit and receive technology can offer a data rate of 0.5 Mbps. A
`two-transmit and one-receive antenna system would achieve 0.8 Mbps. A four-transmit
`and four-receive antenna system can reach 3.75 Mbps. It is worth noting that 3.75 Mbps
`is also achievable in a single antenna transmit and receive technology, but needs 10°
`times higher SNR or transmit power compared with a four-transmit and four-receive
`antenna configuration. The technology that can deliver such remarkable gainsis the
`subject of this book.
`
`1.2._Exploiting multiple antennasin wireless
`
`Figure 1.5 illustrates different antenna configurations for ST wireless links. SISO (sin-
`gle input single output) is the familiar wireless configuration, SIMO (single input
`multiple output) has a single transmit antenna and multiple (M,) receive antennas,
`MISO (multiple input single output) has multiple (M7) transmit antennas and a sin-
`gle receive antenna and MIMO(multiple input multiple output) has multiple (M7)
`
`IPR2018-01477
`Apple Inc. EX1009 Page 17
`
`IPR2018-01477
`Apple Inc. EX1009 Page 17
`
`
`
`1.2 Exploiting multiple antennas in wireless
`7
`aeee
`
`transmit antennas and multiple (7p) receive antennas. The MIMO-MU (MIMO mul-
`tiuser) configuration refers to the case where a base-station with multiple (/) antennas
`communicates with P users each with one or more antennas. Both transmit and re-
`ceive configurations are shown. We sometimes abbreviate SIMO, MISO and MIMO
`configurations as XIXO.
`
`1.2.1
`
`Array gain
`
`Array gain refers to the average increase in the SNRatthe receiverthat arises from the
`coherent combining effect of multiple antennasat the receiver or transmitter or both.
`Consider, as an example, a SIMO channel. Signals arriving at the receive antennas
`have different amplitudes and phases. The receiver can combinethesignals coherently
`so that the resultant signal is enhanced. The average increase in signal powerat the
`receiver is proportional to the number of receive antennas. In channels with multiple
`antennasat the transmitter (MISO or MIMO channels), array gain exploitation requires
`channel knowledgeat the transmitter.
`
`1.2.2
`
`Diversity gain
`
`Signal powerin a wireless channel fluctuates (or fades). When the signal power drops
`significantly, the channelis said to be in a fade. Diversity is used in wireless channels
`to combatfading.
`Receive antennadiversity can be used in SIMO channels [Jakes, 1974]. The receive
`antennas see independently faded versions of the same signal. The receiver combines
`these signals so that the resultant signal exhibits considerably reduced amplitude vari-
`ability (fading) in comparison with the signal at any one antenna. Diversity is charac-
`terized by the number of independently fading branches, also knownasthe diversity
`order and is equal to the numberof receive antennas in SIMO channels.
`Transmit diversity is applicable to MISO channels and has becomean active area for
`- research [Wittneben, 1991; Seshadri and Winters, 1994; Kuo andFitz, 1997; Olofsson
`et al., 1997; Heath and Paulraj, 1999]. Extracting diversity in such channels is possible
`with or without channel knowledge atthe transmitter. Suitable design of the transmitted
`signal is required to extract diversity. ST diversity coding [Seshadri and Winters, 1994;
`Gueyer al., 1996; Alamouti, 1998; Tarokh ef a/., 1998, 1999b] is a transmitdiversity
`technique that relies on coding across space (transmit antennas) to extract diversity
`in the absence of channel knowledgeat the transmitter. If the channels of all transmit
`antennas to the receive antenna have independent fades, the diversity order of this
`channelis equal to the numberof transmit antennas.
`Utilization of diversity in MIMO channels requires a combination of the receive and
`transmit diversity described above. The diversity order is equal to the product of the
`
`IPR2018-01477
`Apple Inc. EX1009 Page 18
`
`IPR2018-01477
`Apple Inc. EX1009 Page 18
`
`
`
`Introduction
`1
`
`
`numberof transmit and receive antennas, if the channel between each transmit—receive
`antenna pair fades independently.
`
`1.2.3
`
`Spatial multiplexing (SM)
`
`SM offers a linear (in the numberof transmit—receive antennapairs or min(Mr, Mr))
`increase in the transmission rate (or capacity) for the same bandwidth and with no
`additional power expenditure. SM is only possible in MIMO channels [Paulraj and
`Kailath, 1994; Foschini, 1996; Telatar, 1999a]. In the following we discuss the basic
`principles of SM for a system with two transmit and two receive antennas. The concept
`can be extended to more general MIMOchannels.
`The bit stream to be transmitted is demultiplexed into two half-rate sub-streams,
`modulated and transmitted simultaneously from each transmit antenna. Underfavor-
`able channel conditions, the spatial signatures of these signals induced at the receive
`antennas are well separated. The receiver, having knowledge of the channel, candif-
`ferentiate between the two co-channel signals and extract both signals, after which
`demodulation yields the original sub-streams that can now be combinedto yield the
`original bit stream. Thus SM increasestransmission rate proportionally with the number
`of transmit-receive antennapairs.
`SM can also be applied in a multiuser format (MIMO-MU, also known as space
`division multiple access or SDMA). Consider two users transmitting their individual
`signals, which arrive at a base-station equipped with two antennas. The base-station
`can separate the two signals to support simultaneoususe of the channel by both users.
`Likewise the base-station can transmit two signals with spatial filtering so that each
`user can decode its ownsignal adequately. This allows a capacity increase proportional
`to the numberofantennasatthe base-station and the numberofusers.
`
`1.2.4
`
`Interference reduction
`
`Co-channelinterference arises due to frequency reuse in wireless channels. When mul-
`tiple antennasare used,the differentiation between the spatial signatures of the desired
`signal and co-channelsignals can be exploited to reducethe interference. Interference
`reduction requires knowledge of the channel of the desired signal. However, exact
`knowledgeof the interferer’s channel may not be necessary.
`Interference reduction (or avoidance) can also be implemented at the transmitter,
`wherethe goalis to minimize the interference energy sent towards the co-channel users
`while delivering the signal to the desired user. Interference reduction allows the use of
`aggressive reuse factors and improves network capacity.
`We note that it may not be possible to exploit all the leverages simultaneously due
`to conflicting demandson the spatial degrees of freedom (or numberof antennas). The
`degree to which these conflicts are resolved depends uponthe signaling scheme and
`receiver design.
`
`IPR2018-01477
`Apple Inc. EX1009 Page 19
`
`Lap
`MN
`
`1m)
`
`IPR2018-01477
`Apple Inc. EX1009 Page 19
`
`
`
`9
`Sees
`
`1.3 ST wireless communication systems
`
`
`
` ST coding
` post-filtering
`
`interleaving
`
`
`Modulation
`RF
`pre-filtering
`
`¥
`
`
`
`SS
`
`Deinterleaving
`ST receiver
`
`
`
`Figure 1.6: Schematic of a ST wireless communication system.
`
`1.3
`
`ST wireless communication systems
`
`Figure 1.6 shows a typical ST wireless system with My transmit antennas and Mz
`receive antennas. The input data bits enter a ST coding block that adds parity bits
`for protection against noise and also captures diversity from the space and possibly
`frequency or time dimensions in a fading environment. After coding,the bits (or words)
`are interleaved across space, time and frequency and mapped to data symbols (such
`as quadrature amplitude modulation (QAM)) to generate M;y outputs. The My; symbol
`streams may then be ST pre-filtered before being modulated with a pulse shaping
`function, translated to the passband via parallel RF chains and then radiated from Mr
`antennas. These signals pass through the radio channel where they are attenuated and
`undergo fading in multiple dimensions before they arrive at the Mp receive antennas.
`Additive thermal noise in the Mz parallel RF chainsat the receiver corrupts the received
`signal. The mixture ofsignal plus noise is matched-filtered and sampled to produce Mr
`output streams. Some form of additional ST post-filtering may also be applied. These
`streams are then ST deinterleaved and ST decoded to producethe outputdata bits.
`The difference between a ST communication system and a conventional system
`comesfrom the use of multiple antennas, ST encoding/interleaving, ST pre-filtering
`and post-filtering and ST decoding/deinterleaving.
`We concludethis chapter with a brief overviewof the areas discussed in the remainder
`of this book. Chapter 2 overviews ST propagation. We develop a channel representation
`as a vector valued ST random field and derive multiple representations andstatistical
`descriptions of ST channels. We also describe real world channel measurements and
`models.
`Chapter 3 introduces XIXO channels, derives channels from statistical ST channel
`descriptions, proposes general XIXO channel models andtest channel models and ends
`with a discussion on XIXO channelestimation at the receiver and transmitter.
`Chapter 4 studies channel capacity of XIXO channels under a variety of conditions:
`channel knownand unknownto thetransmitter, general channel models and frequency
`
`IPR2018-01477
`Apple Inc. EX1009 Page 20
`
`IPR2018-01477
`Apple Inc. EX1009 Page 20
`
`
`
`10
`
`Introduction
`1
`
`
`selective channels. We also discuss the ergodic and outage capacity of random XIXO
`channels.
`Chapter 5 overviews the spatial diversity for XIXO channels, bit error rate perfor-
`mance with diversity and the influence of general channel conditions on diversity and
`ends with techniques that can transformspatial diversity at the transmitter into time or
`frequency diversity at the receiver.
`Chapter 6 develops ST coding for diversity, SM and hybrid schemes for single carrier
`modulation where the channel is not knownat the transmitter. We discuss performance
`criteria in frequency flat and frequency selective fading environments.
`Chapter 7 describes ST receivers for XIXO channels and for single carrier modula-
`tion. We discuss maximum likelihood (ML), zero forcing (ZF), minimum mean square
`error (MMSE) and successive cancellation (SUC) receiver structures. Performance
`analysis is also provided.
`Chapter 8 addre