EE 5407 Part II: Spatial Based Wireless Communications

Size: px

Start display at page:

Download "EE 5407 Part II: Spatial Based Wireless Communications"

Harry Lynch
6 years ago
Views:

1 EE 5407 Part II: Spatial Based Wireless Communications Instructor: Prof. Rui Zhang Website: Lecture I: Introduction March 4,

2 About The Instructor Senior Research Engineer, Institute for Infocomm Research (I2R); Assistant Professor (joint appointment), ECE Department, NUS B. ENG & M. ENG, ECE Department, NUS; Ph.D, EE Department, Stanford University Current Research Interests: (1) Wireless Communication (Multiuser MIMO, Cognitive Radio, Cooperation Communication, Energy Efficiency & Energy Harvesting); (2) Convex Optimization for Applications in Signal Processing & Communication; (3) Information Theory 2

3 Outline Overview of Wireless Communications Introduction to Multi-Antenna Wireless Communications Course Overview 3

4 Overview of Communications Systems There are seven layers (physical, data link, network, transport, session, presentation, application) in the Open Systems Interconnection (OSI) model. The three lower layers are closely related to communication system design. Physical Layer: Transmitter Channel Receiver Transmitter: channel coding, modulation, pulse shaping, power control, precoding,... Channel: impairments due to path loss, multipath propagation, time variation of the channel, interference and noise Receiver: equalization, demodulation, channel decoding,... Data-Link Layer: Medium Access Control (MAC): Protocols to allow frames to be sent over the shared media without undue interference to other nodes. 4

5 Multiple Access Schemes: Physical layer schemes to support multiple users to communicate with the same base station (access point). Time division multiple access (TDMA) Frequency division multiple access (FDMA) Code division multiple access (CDMA) Spatial division multiple access (SDMA) Orthogonal frequency division multiple access (OFDMA) Network Layer: Routing: Define the route by which the message from the source is passed to the ultimate destination within the network. Quality of Service (QoS) Control: throughput, delay,... 5

6 Overview of Wireless Communication Systems Cellular mobile systems Most popular To provide wireless connections anytime, anywhere Voice dominated; video transmission and high speed internet are also in high demand Mobility requirement Range: kilometer to tens of kilometers Standardization activities and multiple access schemes 2G/2.5G: GSM TDMA; IS-95 CDMA 3G: CDMA-2000, WCDMA, TDS-CDMA 3.5G: HSDPA, HSUPA 3.9G (3G Long Term Evolution): OFDMA, IFDMA 6

7 4G: OFDMA+SDMA (MIMO-OFDM) Challenges: capacity, coverage, high data rate Wireless local area networks (WLAN, WiFi) Providing high speed wireless connections for LAN environment Hot spots (airport, hotel, shopping mall), office and home environment, campus,... Standardization activities Low rate: IEEE802.11b (2-11Mbps, 2.4GHz) Medium rate: IEEE802.11a (6-54Mbps, 5.2GHz), IEEE802.11g (6-54Mbps, 2.4GHz); OFDM High rate: IEEE802.11n (peak rate around 600Mbps, 5.2GHz, thanks to MIMO technology); MIMO-OFDM Range: < 20 m for peak data rate; around 100m for lowest rate Challenges: high data rate, coverage 7

8 Fixed/mobile wireless data services Last-mile broadband wireless access technique Alternative or complement to cable modem, DSL Data-dominated services Range: tens of kilometers Standardization activities IEEE802.16: GHz; Jan 2001 IEEE802.16e (WiMAX): 2-11 GHz; Jan 2006; with mobility; OFDM, SCCP, MIMO. IEEE (Wireless regional area networks (WRAN) based on cognitive radio technology): 54 MHz MHz (TV channels) Challenges: coverage, high data rate Wireless personal area networks (WPAN) Providing high speed wireless connections for very short distance 8

9 using ultra wideband (UWB) technology To replace physical cables, wireless USB,... Home networking Peak data rate: 480Mbps Standardization activity: IEEE working group. Multiband OFDM with bandwidth about 500 MHz Impulse radio based DS-CDMA with bandwidth about 2GHz Range: < 10 m Challenges: coverage 9

10 Impairments for Wireless Transmissions Path loss low received signal-to-noise ratio (SNR) low rate, short distance Shadowing and Fading received SNR fluctuation increased BER Multi-path inter-symbol interference (ISI) low received signal-to-interference-plus-noise ratio (SINR) increased BER Co-channel interference (occurs when two or more users operate over the same frequency band at the same time) low received SINR increased BER, low capacity 10

11 General Challenges For Wireless Communications Given the fact that transmission resources such as power and bandwidth are limited, the challenges are How to increase the channel capacity (data rate) without increasing the bandwidth and transmission power? For a given transmission rate, how to extend the coverage without increasing the transmission power? How to increase the system capacity with multiple users? Answer: Exploiting spatial dimension using multiple antennas 11

12 Potential Gains Achieved By Antenna Arrays Array gain: the increase of average SNR at the receiver that arises from the coherent combining effect of multiple antennas at the receiver or transmitter or both. Diversity gain: the increase of transmission reliability by compensating for channel fading via exploiting spatial channel diversity. Spatial multiplexing gain: the increase of data rate by communicating multiple data streams from multiple transmit antennas to multiple receive antennas (without power or bandwidth increase). Interference mitigation gain: the increase of received SINR by nulling/suppressing the co-channel interference via antenna-array beamforming. 12

13 Multi-Antenna Channel Model Single-User (Point-to-Point) Transmission: 13

14 Signal Model: MIMO Case Consider the narrow-band transmission over the flat-fading channel, which is valid when channel coherence bandwidth is much larger than transmission signal bandwidth or channel multi-path delay spread is much smaller than transmission symbol period Consider the block-fading (slow-fading) channel model channel is constant during each transmission block (consisting of many symbols), but may change from one block to another valid when channel coherence time is much larger than block duration 14

15 Discrete-time baseband signal model for each block transmission: r: number of antennas at the receiver t: number of antennas at the transmitter n: symbol time index y(n) C r 1 : received signal vector y(n) = Hx(n) + z(n) (1) H C r t : channel matrix; the element at the ith row and jth column of H, denoted by h ij = [H] i,j, is the complex channel coefficient from the jth transmit antenna to ith receive antenna, i {1,...,r}, j {1,...,t} x(n) C t 1 : transmitted signal vector; it is assumed that x(n) is circularly symmetric random vector, and has zero mean, E[x(n)] = 0, and covariance matrix, S x E[x(n)x H (n)], where S x is Hermitian symmetric, i.e., S x = S H x 15

16 S x is positive semi-definite, i.e., for any vector v C t 1, v H S x v 0. Note that all eigenvalues of S x are non-negative real numbers, and the eigenvalue decomposition of S x can be written by S x = U x Λ x U H x, where U x C t t and U x U H x = I, and Λ x is a t t diagonal matrix with the diagonal elements being the eigenvalues of S x z(n) C r 1 : noise vector at the receiver; it is assumed that z(n) s are independent over n, and z(n) is circularly symmetric and jointly Gaussian random vector, referred to as circularly symmetric complex Gaussian (CSCG), and has zero mean and covariance matrix, S z E[z(n)z H (n)]; for brevity, denote z(n) CN(0, S z ) z(n) is independent of x(n) n 16

17 Circularly Symmetric Distribution Definition: The random vector x is circularly symmetric if e jφ x has the same probability distribution as x for all real φ. Theorem: Assume that z is a complex jointly-gaussian random vector with zero mean. Then z is circularly symmetric if and only if (iif) M z E[zz T ] = 0. Let z = [z 1,..., z N ] T. Note that M z = 0 implies that E[(Re(z i )) 2 ] = E[(Im(z i )) 2 ], i {1,...,N} E[Re(z i ) Im(z i )] = 0, i {1,...,N} E[Re(z i ) Re(z j )] = E[Im(z i ) Im(z j )], i, j {1,...,N}, i j E[Re(z i ) Im(z j )] = E[Im(z i ) Re(z j )], i, j {1,...,N}, i j The distribution or probability density function (PDF) of a zero-mean CSCG random vector z depends only on S z. 17

18 MIMO Channel Distribution IID (independent and identically distributed) Rayleigh-fading channel: H consists of independent CSCG random variables each with zero mean and variance σ 2, i.e., h ij CN(0, σ 2 ), i, j. Corresponds to rich-scattering environments at both transmitter and receiver sides In this case, we can write h ij = α ij e jθ ij, where θ ij = h ij is uniformly distributed over [0, 2π); and α ij = h ij is Rayleigh distributed with PDF: f α (x) = 2x x 2 σ 2 e Then β ij = α 2 ij is exponentially distributed with PDF: σ 2 (2) f β (y) = 1 σ 2e y σ 2 (3) 18

19 For notational brevity, denote H H w MIMO channel with transmit/receive antenna correlations: If the antennas are correlated at the transmitter side, but not at the receiver side, the MIMO channel can be modeled by H = H w R 1/2 t (4) where R t C t t is the covariance matrix describing the transmit antenna correlations. Similarly, the MIMO channel with the correlated receive antennas and uncorrelated transmit antennas can be modeled by H = R 1/2 r H w (5) where R r C r r is the covariance matrix describing the receive antenna correlations. Generally, in the presence of both transmit and receive antenna 19

20 correlations, the MIMO channel can be modeled by H = R 1/2 r H w R 1/2 t. (6) The IID channel matrix H w is a full-rank matrix with probability 1, i.e., the rank of H w satisfies Rank(H w ) = min(t, r). In the presence of transmit and receive antenna correlations, Rank(H) min(rank(r r ),Rank(R t )), with probability 1. There are many other MIMO channel models: Rician fading (with LOS component), Degenerate channels (e.g., pin-hole channel),... 20

21 Frequency Selective Fading MIMO Channel Consider broadband transmission over frequency selective fading channel applicable when signal bandwidth (inverse of symbol duration) is comparable with channel coherence bandwidth (inverse of multi-path delay spread), and thus two or more signal propagation paths are resolvable at the receiver Consider the slow-fading/block-fading channel model Discrete-time baseband signal model for each block transmission: y(n) = L 1 l=0 L 1: number of resolvable multi-paths H l x(n l) + z(n) (7) H l C r t : channel matrix for the lth path, l {0,...,L 1} 21

22 Multiuser MIMO System Model 22

23 Spatial Division Multiple Access (SDMA) Two or more users each with multiple transmit and/or receive antennas communicate over the same frequency band and at the same time slot (e.g., in cellular systems) maximizes system capacity creates co-channel interference: spatial interference mitigation is needed (a very active area of research) MAC (SIMO, MIMO): models the uplink (UL) transmission of a single cell; independent transmit processing and joint receive processing BC (MISO, MIMO): models the downlink (DL) transmission of a single cell; joint transmit processing and independent receive processing IC (MISO, SIMO, MIMO): models the UL/DL transmission of two or more cells; independent transmit/receive processing 23

24 Course Overview Introduction (Lecture I) Overview of multi-antenna wireless communications MIMO channel and signal models Receive Beamforming (Lecture II) SIMO channel Receive beamforming techniques: selection combining, equal-gain combining, maximal-ratio combining Diversity order, array gain Transmit Beamforming & Transmit Diversity (Lecture III) MISO channel Transmit beamforming with (w/) Channel State Information at 24

25 Transmitter (CSIT) Transmit diversity without (w/o) CSIT: Alamouti code MIMO channel: Joint transmit beamforming and receive beamforming w/ CSIT Joint transmit diversity and receive beamforming w/o CSIT MIMO Systems (Lecture IV) Overview of single-antenna/siso AWGN (additive white Gaussian noise) and fading channel capacities MIMO AWGN channel: capacity, transceiver design for spatial multiplexing (CSIT-known vs. CSIT-unknown), MIMO detection MIMO fading channel: ergodic capacity, outage capacity MIMO-OFDM (Lecture V) OFDM for SISO frequency selective fading channel OFDM for MIMO frequency selective fading channel: MIMO-OFDM 25

26 Course Logistics In total, six lectures Two continuous assessments (CAs): each counts 10% in final grade Due date for CA I: March 30, 2011 (in class, firm) Due data for CA II: April 13, 2011 (in class, firm) Final exam (counts 30% in final grade) No tutorials Bonus marks: Detected typo: 0.5 mark each Detected technical error: 1 mark each Total marks capped by 5 (out of 100 in final grade) 26

27 Reference Books A. J. Goldsmith, Wireless Communications, Cambridge University Press, D. Tse and P. Viswanath, Fundamentals of Wireless Communication, Cambridge University Press, A. Paulraj, R. Nabar, and D. Gore, Introduction to space-time wireless communications, Cambridge University Press,

CHAPTER 8 MIMO. Xijun Wang

CHAPTER 8 MIMO. Xijun Wang CHAPTER 8 MIMO Xijun Wang WEEKLY READING 1. Goldsmith, Wireless Communications, Chapters 10 2. Tse, Fundamentals of Wireless Communication, Chapter 7-10 2 MIMO 3 BENEFITS OF MIMO n Array gain The increase