Capacity Region of the Two-Way Multi-Antenna Relay Channel with Analog Tx-Rx Beamforming

Similar documents
Exploiting Partial Channel Knowledge at the Transmitter in MISO and MIMO Wireless

ELEC E7210: Communication Theory. Lecture 10: MIMO systems

Single-User MIMO systems: Introduction, capacity results, and MIMO beamforming

Half-Duplex Gaussian Relay Networks with Interference Processing Relays

Degrees-of-Freedom for the 4-User SISO Interference Channel with Improper Signaling

Multi-User Gain Maximum Eigenmode Beamforming, and IDMA. Peng Wang and Li Ping City University of Hong Kong

Trust Degree Based Beamforming for Multi-Antenna Cooperative Communication Systems

Incremental Coding over MIMO Channels

TO exploit the benefits (e.g., diversity or multiplexing. A General Criterion for Analog Tx-Rx Beamforming under OFDM Transmissions

12.4 Known Channel (Water-Filling Solution)

Multi-Input Multi-Output Systems (MIMO) Channel Model for MIMO MIMO Decoding MIMO Gains Multi-User MIMO Systems

Lecture 5: Antenna Diversity and MIMO Capacity Theoretical Foundations of Wireless Communications 1. Overview. CommTh/EES/KTH

Generalized Signal Alignment: On the Achievable DoF for Multi-User MIMO Two-Way Relay Channels

Physical-Layer MIMO Relaying

EE 5407 Part II: Spatial Based Wireless Communications

Optimal Beamforming for Two-Way Multi-Antenna Relay Channel with Analogue Network Coding

Optimal Transmit Strategies in MIMO Ricean Channels with MMSE Receiver

Novel spectrum sensing schemes for Cognitive Radio Networks

On the Optimization of Two-way AF MIMO Relay Channel with Beamforming

Limited Feedback in Wireless Communication Systems

Cooperative Communication in Spatially Modulated MIMO systems

This document is downloaded from DR-NTU, Nanyang Technological University Library, Singapore.

DEVICE-TO-DEVICE COMMUNICATIONS: THE PHYSICAL LAYER SECURITY ADVANTAGE

Anatoly Khina. Joint work with: Uri Erez, Ayal Hitron, Idan Livni TAU Yuval Kochman HUJI Gregory W. Wornell MIT

ABSTRACT. Cooperative Detection and Network Coding in Wireless Networks. Mohammed W. Baidas, Master of Science, 2009

Vector Channel Capacity with Quantized Feedback

Under sum power constraint, the capacity of MIMO channels

Degrees-of-Freedom Robust Transmission for the K-user Distributed Broadcast Channel

ELG7177: MIMO Comunications. Lecture 8

Multi-Input Multi-Output Systems (MIMO) Channel Model for MIMO MIMO Decoding MIMO Gains Multi-User MIMO Systems

POWER ALLOCATION AND OPTIMAL TX/RX STRUCTURES FOR MIMO SYSTEMS

On the Optimality of Multiuser Zero-Forcing Precoding in MIMO Broadcast Channels

Lecture 8: MIMO Architectures (II) Theoretical Foundations of Wireless Communications 1. Overview. Ragnar Thobaben CommTh/EES/KTH

Sum-Rate Maximization in Two-Way AF MIMO Relaying: Polynomial Time Solutions to a Class of DC Programming Problems

Lecture 7 MIMO Communica2ons

The Optimality of Beamforming: A Unified View

Lecture 9: Diversity-Multiplexing Tradeoff Theoretical Foundations of Wireless Communications 1

Improper Gaussian signaling for

Lecture 9: Diversity-Multiplexing Tradeoff Theoretical Foundations of Wireless Communications 1. Overview. Ragnar Thobaben CommTh/EES/KTH

When does vectored Multiple Access Channels (MAC) optimal power allocation converge to an FDMA solution?

The Effect of Channel State Information on Optimum Energy Allocation and Energy Efficiency of Cooperative Wireless Transmission Systems

The Effect of Spatial Correlations on MIMO Capacity: A (not so) Large N Analytical Approach: Aris Moustakas 1, Steven Simon 1 & Anirvan Sengupta 1,2

ROBUST BEAMFORMING FOR OFDM MODULATED TWO-WAY MIMO RELAY NETWORK. A Thesis JIANWEI ZHOU

Approximate Ergodic Capacity of a Class of Fading Networks

High SNR Analysis for MIMO Broadcast Channels: Dirty Paper Coding vs. Linear Precoding

A Systematic Approach for Interference Alignment in CSIT-less Relay-Aided X-Networks

Signaling Design of Two-Way MIMO Full-Duplex Channel: Optimality Under Imperfect Transmit Front-End Chain

Performance Analysis of MIMO Relay Networks with Beamforming. Hyunjun Kim

Secure Degrees of Freedom of the MIMO Multiple Access Wiretap Channel

Cooperative Interference Alignment for the Multiple Access Channel

Advanced Topics in Digital Communications Spezielle Methoden der digitalen Datenübertragung

Comparisons of Performance of Various Transmission Schemes of MIMO System Operating under Rician Channel Conditions

Capacity optimization for Rician correlated MIMO wireless channels

Multiuser Downlink Beamforming: Rank-Constrained SDP

On the Degrees of Freedom of the Finite State Compound MISO Broadcast Channel

Lecture 6: Modeling of MIMO Channels Theoretical Foundations of Wireless Communications 1

Lecture 6: Modeling of MIMO Channels Theoretical Foundations of Wireless Communications 1. Overview. CommTh/EES/KTH

Homework 5 Solutions. Problem 1

Transmit Directions and Optimality of Beamforming in MIMO-MAC with Partial CSI at the Transmitters 1

Ergodic and Outage Capacity of Narrowband MIMO Gaussian Channels

Multiple Antennas in Wireless Communications

Degrees of freedom of wireless interference network

ELEC546 MIMO Channel Capacity

Degrees of Freedom Region of the Gaussian MIMO Broadcast Channel with Common and Private Messages

Multiple Antennas for MIMO Communications - Basic Theory

Joint FEC Encoder and Linear Precoder Design for MIMO Systems with Antenna Correlation

MIMO Multiway Relaying with Clustered Full Data Exchange: Signal Space Alignment and Degrees of Freedom

Lecture 2. Capacity of the Gaussian channel

SOS-BASED BLIND CHANNEL ESTIMATION IN MULTIUSER SPACE-TIME BLOCK CODED SYSTEMS

Performance Analysis of MIMO-OSTBC based Selective DF Cooperative Wireless System with Node Mobility and Channel Estimation Errors

Dirty Paper Coding vs. TDMA for MIMO Broadcast Channels

Morning Session Capacity-based Power Control. Department of Electrical and Computer Engineering University of Maryland

Spectrum Leasing via Cooperation for Enhanced. Physical-Layer Secrecy

Random Access Protocols for Massive MIMO

Semi-Definite Programming (SDP) Relaxation Based Semi-Blind Channel Estimation for Frequency-Selective MIMO MC-CDMA Systems

Achievable Outage Rate Regions for the MISO Interference Channel

Secrecy Outage Performance of Cooperative Relay Network With Diversity Combining

L interférence dans les réseaux non filaires

Simultaneous SDR Optimality via a Joint Matrix Decomp.

Interactive Interference Alignment

Outline - Part III: Co-Channel Interference

Multi-Antenna Cooperative Wireless Systems: A Diversity-Multiplexing Tradeoff Perspective

Mode Selection for Multi-Antenna Broadcast Channels

Sub-modularity and Antenna Selection in MIMO systems

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 15, NO. 11, NOVEMBER

Cooperative Transmission for Wireless Relay Networks Using Limited Feedback

On the Capacity and Degrees of Freedom Regions of MIMO Interference Channels with Limited Receiver Cooperation

An Uplink-Downlink Duality for Cloud Radio Access Network

Pilot Optimization and Channel Estimation for Multiuser Massive MIMO Systems

Optimum Transmission Scheme for a MISO Wireless System with Partial Channel Knowledge and Infinite K factor

IN this paper, we show that the scalar Gaussian multiple-access

COM Optimization for Communications 8. Semidefinite Programming

Ergodic and Outage Capacity of Narrowband MIMO Gaussian Channels

Comparison of Full-Duplex and Half-Duplex Modes with a Fixed Amplify-and-Forward Relay

2318 IEEE TRANSACTIONS ON SIGNAL PROCESSING, VOL. 54, NO. 6, JUNE Mai Vu, Student Member, IEEE, and Arogyaswami Paulraj, Fellow, IEEE

Optimized Beamforming and Backhaul Compression for Uplink MIMO Cloud Radio Access Networks

Optimization in Wireless Communication

SINR Balancing in the Downlink of Cognitive Radio Networks with Imperfect Channel Knowledge

Clean relaying aided cognitive radio under the coexistence constraint

Schur-convexity of the Symbol Error Rate in Correlated MIMO Systems with Precoding and Space-time Coding

Transcription:

Capacity Region of the Two-Way Multi-Antenna Relay Channel with Analog Tx-Rx Beamforming Authors: Christian Lameiro, Alfredo Nazábal, Fouad Gholam, Javier Vía and Ignacio Santamaría University of Cantabria, Santander, SPAIN

Outline 1 RF-MIMO Two-Way Relay Channel (RF-TWRC) 2 Capacity Region 3 Semidefinite Relaxation 4 Numerical Examples 5 Conclusion

Two-way relay channel Two-phase protocol......... 1 R 2 Multiple access channel (MAC).......... 1 R 2 Broadcast channel (BC). Amplify-and-forward (AF) strategy with two-phase protocol. MAC phase: the nodes transmit simultaneously to the relay. BC phase: the relay retransmits the linearly processed signal. Perfect CSI: each node is able to null out the interference.

Overview of the state-of-the-art The capacity region of the TWRC-AF when the source nodes are single antenna terminals with fixed powers has been computed (R. Zhang et al., 2009). A suboptimal algorithm has been recently proposed for the MIMO-TWRC with AF strategy (Wang and Zhang, 2010). The capacity region of the TWRC with decode-and-forward strategy has been computed (R.F. Wyrembelski et al., 2009). In this work, we consider the multi-antenna TWRC-AF when the nodes perform analog beamforming.

RF-MIMO terminals based on analog beamforming Reduction of system size, hardware cost and power consumption. Multiplexing gain equal to 1, but full array and diversity gains. Point-to-point links, BC and MAC thoroughly studied.

System model (I) In the MAC phase, each node performs transmit beamforming with {v 1,v 2 } C NS 1, respectively; and the relay receives through the beamformer u R C NR 1. In the BC phase, the relay applies the transmit beamformer v R C NR 1 to the received signal, and the nodes perform receive beamforming with {u 1,u 2 } C NS 1, respectively. The power transmitted by the relay, assuming without loss of generality v R 2 = 1, is p R = p 1 u H R H 1 v 1 2 +p2 u H R H 2 v 2 2 +σ 2 u R 2

System model (II) Assuming perfect CSI, each node removes the self-interference before decoding its desired message. Rx signal: y i = u H i HT i v Ru H R H jv j pj s j + r i, i,j = 1,2,i j H 1 and H 2 are flat fading MIMO channels, and r i is AWGN with zero mean and variance σ [1+ u 2 R 2 u H i H T i v 2] R, i = 1,2. Goal Under power constraints at the nodes and the relay, obtain u 1, u 2, u R, v 1, v 2, v R and the power allocation to operate at any point of the capacity boundary.

Outline 1 RF-MIMO Two-Way Relay Channel (RF-TWRC) 2 Capacity Region 3 Semidefinite Relaxation 4 Numerical Examples 5 Conclusion

Optimal node beamformers Maximum ratio transmission (MRT): v i = HH i u R H H i u, i = 1,2 R Maximum ratio combining (MRC): u i = HT i v R H T i v R, i = 1,2 The optimal relay beamformers lie in the subspace spanned by the columns of the channel matrices, and can be expressed as u R = Ua r H 1 = UG 1 v R = U a t H 2 = UG 2 where U contain the left eigenvectors of [H 1,H 2 ].

The achievable bidirectional rate pairs are R 12 1 2 log p ef1 G T 2 1+ 2 a t 2 σ (1+ a 2 r 2 G T 2 a 2) t R 21 1 2 log p ef2 G 2 1+ T 1 a 2 t ( σ 2 1+ a r 2 ) G T 1 a t 2 where p efi = p i G H i a r 2 is the effective power of node i. Capacity region C(P 1,P 2,P R ) {R 12,R 21 } p 1 P 1,p 2 P 2 a t 2 =1,p R(p 1,p 2,a r) P R

A WSRmax problem cannot be invoked due to its non-convexity. The rate profile method (M. Mohseni et al., 2006) can be used to efficiently characterize the boundary of the capacity region. Proposed algorithm: for a fixed 0 α 1, bisection method over γ sum, solving the following optimization problem in each step. 3 G H 2 G H 2 minimize p 1 1 a r +p2 2 a r +σ 2 a r 2 p 1,p 2,a t,a r 2.5 subject to : p 2 G H 2 a r 2 G T 1 a t 2 σ 2 ( 1+ a r 2 G T 1 at 2 ) αγsum p 1 G H 1 a r 2 G T 2 a t 2 σ 2 ( 1+ a r 2 G T 2 at 2 ) (1 α)γsum R 2 (bps/hz) 2 1.5 1 0.5 Capacity region Rate profile p 1 P 1 0 0.5 1 1.5 2 2.5 3 p 2 P 2 R 1 (bps/hz) 0

Outline 1 RF-MIMO Two-Way Relay Channel (RF-TWRC) 2 Capacity Region 3 Semidefinite Relaxation 4 Numerical Examples 5 Conclusion

The initial problem is non-convex, but a solution can be found through a relaxed semidefinite programm (SDP). New optimization variables Ar = a r a H r Equivalent problem: A t = a t a H t minimize p 1 Tr(R 1 A r )+p 2 Tr(R 2 A r )+σ 2 Tr(A r ) p 1,p 2,A t,a r subject to : p 2 Tr(R 2 A r ) (1 α)γ sum σ 2 Tr(A r ) (1 α)γ sumσ 2 p 1 Tr(R 1 A r ) αγ sum σ 2 Tr(A r ) αγ sumσ 2 Tr(A t ) = 1 A t 0, rank(a t ) = 1, p 1 P 1 A r 0, rank(a r ) = 1, p 2 P 2 where R i = G i G H i, i = 1,2. Tr(R 2 A t) Tr(R 1 A t)

Analysis of the equivalent problem The equivalent problem is still non-convex due to: The cross products between the powers and the beamforming matrices. The rank-one constraints. Managing the non-convexity of the equivalent problem We can avoid the cross products by optimizing the effective powers instead, and changing the power constraints accordingly. We can find a solution of the equivalent problem relaxing the rank-one constraints, what is called a relaxed SDP.

Final optimization problem (I) Convex problem minimize p ef1,p ef2,a t,a r p ef1 +p ef2 +σ 2 Tr(A r ) subject to : p ef2 (1 α)γ sum σ 2 Tr(A r ) (1 α)γ sumσ 2 p ef1 αγ sum σ 2 Tr(A r ) αγ sumσ 2 Tr(A t ) = 1 A t 0 A r 0 p ef1 P 1 Tr(R 1 A r ) p ef2 P 2 Tr(R 2 A r ) Tr(R 2 A t) Tr(R 1 A t)

Final optimization problem (II) Key observation: if the rank of the optimal beamforming matrices is greater than one, we are able to find an optimal rank-one solution through the matrix decomposition theorem for Hermitian matrices (Y. Huang and S. Zhang, 2007). Optimal powers After solving the optimization problem, the optimal powers are given by p ( ) 1 = p ( ) 2 = ( Tr ( Tr p ( ) ef 1 R 1 A ( ) r p ( ) ef 2 R 2 A ( ) r ) )

Outline 1 RF-MIMO Two-Way Relay Channel (RF-TWRC) 2 Capacity Region 3 Semidefinite Relaxation 4 Numerical Examples 5 Conclusion

Conventional MIMO vs. analog beamforming Example scenario with single-antenna nodes, i.e., N S = 1, and fixed powers. The relay has N R = 4 antennas, and the SNR is 10 db. 1.4 1.2 Conventional MIMO RF MIMO 1.4 1.2 Conventional MIMO RF MIMO 1 1 R 2 (bps/hz) 0.8 0.6 R 2 (bps/hz) 0.8 0.6 0.4 ρ = 0.1 0.4 ρ = 0.5 0.2 0.2 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 R 1 (bps/hz) 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 R 1 (bps/hz) As ρ increases, i.e., more collinear channels, the capacity gap between analog and conventional beamforming schemes goes to 0.

Capacity region Example scenario with N S = 2 and N R = 4. The SNR is 10 db and the channels have unit variance. 3 2.5 maximum power transmission 2 R 2 (bps/hz) 1.5 1 RF TWRC SISO TWRC SISO TWRC without power optimization 0.5 0 0 0.5 1 1.5 2 2.5 3 R 1 (bps/hz) Some points of the boundary are achieved when the nodes do not transmit at maximum power.

Sum rate vs. SNR analysis The figure shows the sum-rate capacity through Monte Carlo simulations, considering N S = 1, N R = 4 and fixed powers. 7 6 RF MIMO Conventional MIMO SISO 5 Sum Rate (bps/hz) 4 3 2 1 0 0 2 4 6 8 10 12 14 16 18 20 SNR (db)

Outline 1 RF-MIMO Two-Way Relay Channel (RF-TWRC) 2 Capacity Region 3 Semidefinite Relaxation 4 Numerical Examples 5 Conclusion

Conclusion RF-MIMO wireless radios result in low-cost systems with reduced power consumption. The capacity region of the RF-TWRC has been completely characterized. The optimal beamforming vectors and the power allocation can be efficiently computed using convex optimization techniques. The capacity gap between analog and conventional beamforming schemes, when the nodes are single-antenna, goes towards 0 as the angle between the channels decreases.