Complex orthogonal space-time processing in wireless communications

Transcription

1 University of Wollongong Research Online University of Wollongong Thesis Collection University of Wollongong Thesis Collections 2006 Complex orthogonal space-time processing in wireless communications Le Chung Tran SECTE, Faculty of Engineering and Information Sciences, University of Wollongong, Australia, Recommended Citation Tran, Le Chung, Complex orthogonal space-time processing in wireless communications, PhD thesis, School of Electrical, Computer and Telecommunications Engineering, University of Wollongong, Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library:

2

3 A THESIS ENTITLED Complex Orthogonal Space-Time Processing In Wireless Communications Submitted to the University of Wollongong in fulfilment of the requirements for the degree of Doctor of Philosophy Le Chung Tran Bachelor of Telecommunication Engineering (First Class Honours) Hanoi University of Communication and Transportation, Vietnam, 1997 Master of Telecommunication Engineering Hanoi University of Technology, Vietnam, 2000 Australia, May 3, 2006

4 To my family and PTH & To my supervisors

5 Abstract Multiple-Input Multiple-Output (MIMO) transmission has recently emerged as one of the most significant technical breakthroughs in modern communication with a chance to resolve the bottleneck of traffic capacity in the future wireless networks. Communication theories show that MIMO systems can provide potentially a very high capacity that, in many cases, grows approximately linearly with the number of antennas. Space-time processing is the main feature of MIMO systems. Space-Time Codes (STCs) are the codes designed for the use in MIMO systems. Among a variety of STCs, orthogonal Space-Time Block Codes (STBCs) possess a much simpler decoding method over other STCs. Because of that, this thesis examines orthogonal STBCs in MIMO systems. Furthermore, Complex Orthogonal STBCs (CO STBCs) are mainly considered in this thesis since they can be used for PSK/QAM modulation schemes, and therefore, are more practical than real STBCs. The thesis starts with the backgrounds on MIMO systems and their capacity, on STBCs, and on some conventional transmission diversity techniques. These backgrounds are essential for the readers to overview the up-to-date scenario of the issues related to this thesis. After reviewing the state of the art of the issues related to this thesis and indicating the gaps in the literature, the thesis proposes three new maximum rate, order-8 CO STBCs. These new CO STBCs are amenable to practical implementations because they allow for a more uniform spread of power among the transmitter antennas, while providing better performance than other conventional codes of order 8 for the same ii

6 Abstract iii peak power per transmitter antenna. Based on the new proposed CO STBCs, multi-modulation schemes (MMSs) are proposed to increase the information transmission rate of those new codes of order 8. Simulation results show that, for the same MMSs and the same peak power per transmitter antenna, the three new codes provide better error performance than the conventional CO STBCs of the same order 8. In addition, the method to evaluate the optimal inter-symbol power allocation in the proposed codes in single modulation as well as in different MMSs for both Additive White Gaussian Noise (AWGN) and flat Rayleigh fading channels is proposed. It turns out that, for some modulation schemes, equal power transmission per symbol time slot is not only optimal from the technical point of view, but also optimal in terms of achieving the best symbol error probability. The MMSs, which increase the information transmission rate of CO STBCs, and the method to examine the optimal power allocation for multi-modulated CO STBCs mentioned here can be easily generalized for CO STBCs of other orders. Constructions of square, maximum rate CO STBCs are well known. However, codes constructed via the known methods include numerous zeros, which impede their practical implementation, especially in high data rate systems. This disadvantage is partially overcome by the three new CO STBCs of order 8 mentioned above. Nevertheless, these codes still contain zeros which are undesirable or the design method is neither general nor easy yet. By modifying the Williamson and the Wallis-Whiteman arrays to apply to complex matrices, we discover two construction methods of square, order-4n CO STBCs from square, order-n codes. Applying the proposed methods, we construct square, maximum rate, order-8 CO STBCs with no zeros, such that the transmitted symbols equally disperse through transmitter antennas. These codes have the advantages that the power is equally transmitted via each transmitter antenna during every symbol time slot and that a lower peak power per transmitter antenna is required to achieve the same bit error rates as in the conventional CO STBCs with zeros. The combination of CO STBCs and a closed loop transmission diversity technique using a feedback loop has received a considerable attention in the literature since it

7 Abstract iv allows us to improve performance of wireless communication channels with coherent detection. The thesis proposes an improved diversity Antenna Selection Technique (AST), referred to as the (N + 1, N; K) AST/STBC scheme, to improve further the performance of such channels. Calculations and simulations show that our technique performs well, especially, when it is combined with the Alamouti code [7]. While the combination between STBCs and a closed loop transmission diversity technique in the case of coherent detection has been intensively considered in the literature, it seems to be missing for the case of differential detection. The thesis thus proposes two ASTs for wireless channels utilizing Differential Space-Time Block Codes (DSTBCs), referred to as the AST/DSTBC schemes. These techniques improve significantly the performance of wireless channels using DSTBCs (with differential detection). The proposed AST/DSTBC schemes work very well in independent, flat Rayleigh fading channels as well as in the case of perfect carrier recovery. Does this conclusion still hold in the case of correlated, flat Rayleigh fading channels or in the case of imperfect carrier recovery? To answer this question, first, we propose here a very general, straightforward algorithm for generation of an arbitrary number of Rayleigh envelopes with either equal or unequal power, in wireless channels either with or without Doppler frequency shift effects. The proposed algorithm can be applied to the case of spatial correlation, such as with antenna arrays in Multiple Input Multiple Output (MIMO) systems, or spectral correlation between the random processes like in Orthogonal Frequency Division Multiplexing (OFDM) systems. It can also be used for generating correlated Rayleigh fading envelopes in either discrete-time instants or a real-time scenario. The proposed algorithm is not only more generalized and more precise, but also overcome all shortcomings of the conventional methods. Based on the proposed algorithm, the performance of our AST/DSTBC techniques proposed for systems utilizing DSTBCs in spatially correlated, flat Rayleigh fading channels is analyzed. Finally, the thesis examines the effect of imperfect carrier phase/frequency recovery at the receiver on the bit error performance of our

8 Abstract v AST/DSTBC schemes. The tolerance of differential detection associated with the proposed ASTs to phase/frequency errors is then analyzed. These analyses show that our ASTs not only work well in independent, flat Rayleigh fading channels as well as in the case of perfect carrier recovery, but also are very robust in correlated, flat Rayleigh fading channels as well as in the case of imperfect carrier recovery. The thesis is concluded with useful recommendations on the issues examined here and with a number of future research directions.

9 Statement of Originality This is to certify that the work described in this thesis is entirely my own, except where due reference is made in the text. No work in this thesis has been submitted for a degree to any other university or institution. Signed Le Chung Tran May 3, 2006 vi

10 Acknowledgments I would like to sincerely thank my supervisors, Asso. Prof. Tadeusz A. Wysocki, Prof. Jennifer Seberry and Prof. Alfred Mertins, who, with their great talents and excellent expertise, have efficiently guided me as well as greatly supported me in a very tough studying way towards the PhD degree. They have not only supported me in specialistic field, but also helped me to cope with difficulties in my daily life. I am also indebted to Dr. Beata J. Wysocki, who has encouraged and helped me to study well. I would like to thank them for the enlightenment gained by our collaborations on papers, book chapters and projects. I am grateful to various colleagues who have enhanced my understandings of the subject, in particular to Assistant Prof. Sarah A. Spence, Dr. T. Xia, and Y. Zhao. I would like to thank my family, especially, my parents and my PTH who have been supporting me to achieve this degree. Without them, I would not have been successful. Last but not least, I would like to take this opportunity to thank all lecturers, officers, assistants and colleagues in the School of Electrical, Computer and Telecommunications Engineering as well as in the Telecommunications and Information Technology Research Center - TITR - who have helped and assisted me to study well during the time in the University of Wollongong, Australia. The School of Electrical, Computer and Telecommunications Engineering, the University of Wollongong and TITR are really my second home with respectable people and with wonderful memories. vii

11 Contents 1 Introduction Background Overview of the Thesis Contributions of the Thesis Publications Based on the Thesis Literature Review Introduction Multiple-Input Multiple-Output Wireless Communications MIMO System Model Capacity of Additive White Gaussian Noise Channels with Fixed Channel Coefficients Capacity of Flat Rayleigh Fading Channels Space-Time Block Codes Real Orthogonal Designs Complex Orthogonal Designs - CODs Transmission Diversity Techniques Classification of Transmission Diversity Techniques Spatial Diversity Combining Methods Transmit Diversity Techniques viii

12 CONTENTS ix 2.5 Research Problems Considered in the Thesis and Conclusion New Square, Complex Orthogonal Space-Time Block Codes for Eight Transmitter Antennas Introduction New Complex Orthogonal Designs of Order Eight Decoding Metrics Choice of Signal Constellations Simulation Results Conclusion Multi-Modulation Schemes to Achieve Higher Data Rate Introduction Two New Complex Orthogonal STBCs For Eight Transmitter Antennas MMSs to Increase the Data Rate Optimal Inter-Symbol Power Allocation in Single Modulation and in MMSs AWGN Channels Flat Rayleigh Fading Channels Simulation Results Conclusion Two Novel Construction Classes for Improved, Square CO STBCs Introduction Definitions and Notations Design Methods Examples of Maximum Rate, Square, Order-8 CO STBCs with No Zero Entries

13 CONTENTS x 5.5 Simulation Results Conclusion Transmitter Diversity Antenna Selection Techniques for MIMO Systems Introduction Improved Antenna Selection Technique for Wireless Channels Utilizing STBCs Theoretical Basis of Antenna Selection in Wireless Channels Using STBCs with Coherent Detection The (N + 1, N; K) AST/STBC Scheme Simulation Results Transmitter Diversity Antenna Selection Techniques for Wireless Channels Utilizing DSTBCs Reviews on DSTBCs Definitions, Notations and Assumptions Basis of Transmitter Antenna Selection for Channels Using DSTBCs The General (M, N; K) AST/DSTBC Scheme for Channels Utilizing DSTBCs The Restricted (M, N; K) AST/DSTBC Scheme Spatial Diversity Order of the Proposed ASTs Simulation Results Discussions and Conclusion Performance of Diversity Antenna Selection Techniques in Imperfect Channels Introduction A Generalized Algorithm for the Generation of Correlated Rayleigh Fading Envelopes

14 CONTENTS xi Shortcomings of Conventional Methods Fading Correlation as Functions of Time Delay and Frequency Separation Fading Correlation as Functions of Spatial Separation in Antenna Arrays Generalized Algorithm to Generate Correlated, Flat Rayleigh Fading Envelopes Generation of Correlated Rayleigh Envelopes in a Real-Time Scenario Simulation Results Performance of Diversity Antenna Selection Techniques in Correlated, Flat Rayleigh Fading Channels Using DSTBCs AST/DSTBC Schemes in Correlated, Flat Rayleigh Fading Channels Simulation Results Effect of Imperfect Carrier Recovery on the Performance of the Diversity Antenna Selection Techniques in Wireless Channels Utilizing DSTBCs Effect of Phase Errors on the Performance of the Proposed Antenna Selection Techniques Simulation Results Conclusion Conclusion Introduction Main Conclusions Recommendations Future Works Bibliography 216

15 CONTENTS xii A Symbol Error Probability of M-ary PSK Signals in Flat Rayleigh Fading Channels 229 B Proof of the Decision Metrics for Unitary DSTBCs 231

16 List of Figures 1.1 History and main milestones of STBCs Structure of the thesis The diagram of MIMO systems Capacity of MIMO systems with one Rx antenna (Multi-Input Single-Output (MISO) systems) in fast or block flat Rayleigh fading channels Capacity of MIMO systems with one Tx antenna (Single-Input Multi-Output (SIMO) systems) in fast or block flat Rayleigh fading channels Capacity of MIMO systems with equal numbers of Tx and Rx antennas in fast or block flat Rayleigh fading channels Space-time block encoding Selection combining method Scanning combining method Conventional baseband MRC technique using two receiver antennas Alamouti code vs. transmission without coding with QPSK modulation and 8 PSK modulation A conventional COD of order eight Code Z Code Z Code Z xiii

17 LIST OF FIGURES xiv 3.5 The performance of code Z 2 compared to that of the conventional code Z 1 in Rayleigh fading channels The performance of code Z 3 compared to that of the conventional code Z 1 in Rayleigh fading channels The performance of code Z 4 compared to that of the conventional code Z 1 in Rayleigh fading channels Two new CO STBCs proposed for eight transmitter antennas QAM signal constellation and bit mapping scheme SER vs. r in single modulation and MMSs depending on γ in AWGN channels SER vs. γ with the inter-symbol power ratio r=2 for C 1, r=4 for C 2 and with the optimal values r opt in AWGN channels SER vs. r in single modulation and MMSs depending on γ in flat Rayleigh fading channels SER vs. γ with the inter-symbol power ratio r=2 for C 1, r=4 for C 2 and with the optimal values r opt in flat Rayleigh fading channels Comparison between the proposed codes and the conventional one in [87] in AWGN channels Bit error performance of the code C 1 with different MMSs in AWGN channels Bit error performance of the code C 2 with different MMSs in AWGN channels Comparison between the proposed codes and the conventional one in [87] in flat Rayleigh fading channels Bit error performance of the code C 1 with different MMSs in flat Rayleigh fading channels Bit error performance of the code C 2 with different MMSsin flat Rayleigh fading channels The performance of the proposed code in (5.32), compared to the conventional code Z 1 and the proposed codes Z 2, Z 3, Z 4 in Chapter

18 LIST OF FIGURES xv 6.1 The diagram of the (N + 1, N; K) AST/STBC scheme The proposed structure of the feedback information for channels using STBCs The flow chart of the proposed (N + 1, N; K) AST/STBC scheme BER vs. SNR for the Alamouti code and the Tarokh code G4 [81] with and without antenna selection Transmission of DSTBCs (a) without and (b) with the antenna selection technique The general (M, N; K) AST/DSTBC scheme for systems using DSTBCs Some examples of the transmitter antenna grouping for (a) the restricted (4,2;K) AST/DSTBC, (b) the restricted (3,2;K) AST/DSTBC and (c) the restricted (5,4;K) AST/DSTBC The Alamouti DSTBC with the general (3,2;1) AST/DSTBC and the restricted (3,2;1) AST/DSTBC schemes Square, order-4, unitary DSTBC with the general (5,4;1) AST/DSTBC and the restricted (5,4;1) AST/DSTBC schemes Model to examine the spatial correlation between transmitter antennas Model of a Rayleigh generator for an individual Rayleigh envelope corresponding to a desired normalized autocorrelation function Model for generating N Rayleigh fading envelopes corresponding to a desired normalized autocorrelation function in a real-time scenario Example of three equal power, spectrally correlated Rayleigh fading envelopes with GSM specifications Example of three equal power, spatially correlated Rayleigh fading envelopes with GSM specifications Example of three equal power, spectrally correlated Rayleigh fading envelopes with IEEE a (OFDM) specifications Example of three equal power, spatially correlated Rayleigh fading envelopes with a not positive semi-definite covariance matrix

19 LIST OF FIGURES xvi 7.8 Histograms of Rayleigh fading envelopes produced by the proposed algorithm in the four examples along with a Rayleigh PDF where = σ g 2 j 7.9 Computational effort comparison between the method in [73] and the proposed algorithm Proposed (3,2;1) AST/DSTBC schemes in correlated Rayleigh fading channels corresponding to the set ˆR 1 of covariance matrices Proposed (3,2;1) AST/DSTBC schemes in correlated Rayleigh fading channels corresponding to the set ˆR 2 of covariance matrices Proposed (4,2;1) AST/DSTBC schemes in correlated Rayleigh fading channels corresponding to the set ˆR 1 of covariance matrices Proposed (4,2;1) AST/DSTBC schemes in correlated Rayleigh fading channels corresponding to the set ˆR 2 of covariance matrices The effect of imperfect phase recovery on the performance of the Alamouti DSTBC without our ASTs The effect of imperfect phase recovery on the performance of the general (3,2;1) AST/DSTBC scheme The effect of imperfect phase recovery on the performance of the general (4,2;1) AST/DSTBC scheme

20 List of Tables 2.1 Some typical values of p min (or A(1, n)) for the full-rate, real STBCs The maximum number of variables and the maximum rates of square, real STBCs The maximum number of variables and the maximum rates of square CO STBCs The maximum possible rates of non-square CO STBCs The maximum number of variables of non-square CO STBCs The optimal delay of non-square CO STBCs with the maximum possible rates Normalized channel capacity for several values of transmitter and receiver antenna numbers Number of variables in an amicable pair with n = 8 [35] Decision metrics for decoding code Z Decision metrics for decoding code Z Decision metrics for decoding code Z Decision metrics for decoding code Z Mapping rules for the symbols s 1 and s Mapping rules for the symbols s 3 and s The optimality of power allocation in single modulation and MMSs in AWGN channels xvii

21 LIST OF TABLES xviii 4.4 The optimality of power allocation in single modulation and MMSs in flat Rayleigh fading channels The average processing time reduction of the proposed (N +1, N; K) AST/STBC technique Comparison between the proposed (N + 1, N; K) AST/STBC and the technique proposed in [51] SNR gains (db) of the general (3,2;1) AST/DSTBC and the restricted (3,2;1) AST/DSTBC in the channel using Alamouti DSTBC SNR gains (db) of the proposed (5,4;1) AST/DSTBC schemes in the channel using square, order-4, unitary DSTBC

22 List of Abbreviations 3G 3GPP AOD AST AWGN BER BLAST BPSK BS CDMA CO STBC COD const CSI DPCCH DPSK DS-SS DSTBC DSTM e.g. ECK EGC etc. FEC FH-SS Third Generation wireless technology The Third Generation Partnership Project Amicable Orthogonal Design Antenna Selection Technique Additive White Gaussian Noise Bit Error Rate Bell Lab Layered Space-Time Binary Phase Shift Keying Base Station Code Division Multiple Access Complex Orthogonal Space-Time Block Code Complex Orthogonal Design constant Channel State Information Dedicated Physical Control Channel Differential Phase Shift Keying Direct Sequence Spread Spectrum Differential Space-Time Block Code Differential Space-Time Modulation exempli gratia Exact Channel Knowledge Equal Gain Combining et cetera Forward Error Correction Frequency Hoping Spread Spectrum xix

23 List of Abbreviations xx GCOD GSM i.e i.i.d. IDFT iff ISI LAN LOS LST M-ary M-PSK MC-SS MIMO MISO ML MMS MRC MS OFDM PAM PCU PDF PSK QAM QPSK rms Rx antenna SC SCK SER SIMO Generalized Complex Orthogonal Design Global System for Mobile Communications id est identically independently distributed Inverse Discrete Fourier Transform if and only if Inter-Symbol Interference Local Area Network Line Of Sight Layered Space-Time Code Multiple Level Modulation M-ary Phase Shift Keying Multi-Carrier Spread Spectrum Multiple Input Multiple Output Multi-Input Single-Output Maximum Likelihood Multi-Modulation Scheme Maximum Ratio Combining Mobile Station Orthogonal Frequency Division Multiplexing Pulse Amplitude Modulation Per Channel Use Probability Density Function Phase Shift Keying Quadrature Amplitude Modulation Quadrature Phase Shift Keying root-mean-square Receiver antenna Scanning Combining Statistical Channel Knowledge Symbol Error Rate Single-Input Multi-Output

24 List of Abbreviations xxi SNR Signal-to-Noise Ratio STBC Space-Time Block Code STC Space-Time Code STS Symbol Time Slot STTC Space-Time Trellis Code SVD Singular Value Decomposition Tx antenna Transmitter antenna ULA Uniform Linear Array w. r. t. with respect to WCDMA Wideband Code Division Multiple Access