Research On Precoding Algorithm For MIMO Two-way Relay Systems

Multiple-input-multiple-output (MIMO) technique can provide spatialmultiplexing and diversity gains, to efficiently improve the system capacity and protecttransmission against errors caused by channel fading. For these advantages, MIMO hasbeen widely applied and is served as an indispensable component among the nextgeneration wireless communication system. Meanwhile, the cooperation mechanismemployed by cellular system can expand the network coverage considerably, as well asimprove the quality of service (QoS) for edge users. The multi-antenna two-way relaysystem is the combination of MIMO and cooperation, it can ensure the reliability ofcommunication and increase the transmission rate to a great extent. This paper focus onthe relay optimization and joint optimization issues of the two-way relay system, andthe main contributions are summarized as follows:1We propose a reasonable model for the two-way relay system and analyze thecapacity performance. The system throughput upper bound based onamplification-and-forwarding (AF) protocol is derived.2A novel unconstrained convex optimization algorithm is developed for thetwo-way relay system. The algorithm transforms the power restriction objectivefunction to an unconstrained optimization issue and further obtain the sub-optimalsolutions by employing the relaxation operation. The simulation results indicate that thisalgorithm can reduce the computational complexity and improve the system capacitysignificantly.3We propose an algorithm based on the maximization of F-norm. By setting thecommunication system analog SNR as weighting parameters to execute iterativeoptimization, it obtains the relay precoding matrix. The simulation results show that thealgorithm can improve performance of system sum-rate greatly, and maintain thecomputational complexity in a low level.4Motivated by the joint precoding scheduling, we fix the precoding matrix ofsource (or relay), and optimize the precoding matrix of relay (or source). The iterativemethod is applied to optimize the joint precoding matrix. The simulation results indicatethat the joint source-and-relay optimization mechanism can effectively improve thesystem capacity and largely decrease the computational complexity.