Optimization In A Cooperative MIMO Relay System

Multiple inputs and multiple outputs (MIMO) relay cooperative transmission, as an efficient solution for high data rate and ubiquitous coverage, is a promising key technology for next generation wireless system. The performance of MIMO relay network is dependent on the signal processing approaches adopted by relay nodes. Considering the MIMO Relay system, this thesis focus on signal processing methods and power allocation schemes utilized in multi-antenna relays by optimization theory for the target to propose the best transceiver design to enhance relay nodes.Firstly, in order to understand the relay channel comprehensively, the statistical properties of relay channel are studied under Rician fading distribution, including the distribution of received signals envelope, channel time-domain correlations, and antenna array correlations. It is also given for the correlation of different antenna array architectures.Following the studies of optimal linear receiver structure for relay link and diversity relay link, the relay filters under colored noise are designed based on matrix decomposition. A uniform filter structure is proposed, which, resorting to noise whiting technology, extends the application of Amplify and Forward (AF), Zero forcing (ZF), QR and Singular Value Decomposition (SVD) methods. A new approach of Minimum Mean Square Error (MMSE) filter is also suggested and shown with better performance as QR filters by simulation. Furthermore, started from optimization, a simple and unified mathematic model is established for various optimization objectives in design the relay filters. Relying on the Majorization theory, a uniform framework for optimization of relay filters is given, and a general filter configuration is proposed, which is also extend to ZF receiver and power allocation problem with QoS constraints.Based on these results, the joint optimization of the precoders of source node and relay nodes is started at the next step. Note that similar uniform framework is also found for those problems of joint optimization. In order to solve the joint power allocation of source and relay nodes, an iterative optimal algorithm and two sub- optimal algorithms are proposed and verified by simulation.It is more complex for design of relay filters when multiple users are active in MIMO relay network. Considering the minimum MSE problem, an algorithm is given based on the downlink-uplink duality of a MIMO relay system, and. Moreover, two sub-optimal algorithms are proposed based on the properties of relay filters. The simulation results the sub- optimal algorithms give similar results as optimal one.