Joint Optimization for Source and Relay Precoding under Multiuser MIMO Downlink Channels

This paper investigates the joint precoding optimization problem for a relay-assisted multi-antenna downlink system. Aiming at multiuser sum capacity maximization, we first propose an iterative optimization algorithm which exploits quadratic programming approaches. We observe that the iterative method always yields the optimized precoding matrices which diagonalize the compound channel of the system at high SNR regimes. Inspired by this observation, we further develop an efficient source and relay precoding strategy by diagonalizing the compound channel of the backward (source-to-relay) and the forward (relay-to-destination) links. Simulation results verify the effectiveness of our proposed precoding schemes.