Research On Beamforming Techniques In MIMO Relay Networks

This article studies MIMO, relay and beamforming technologies inthe application of wireless cellular communication system. The model ofmultiple relays or multiple users is considered in the analysis and design.We first consider a circumstance of single user, then a circumstance ofmultiple user, and finally a condition of user selection due to the largenumber of users. For the relay nodes, it is supposed that all the nodes arehalf-duplex and in AF(Amplify-and-Forward) mode, and the condition thatthere is no direct link between the base and the user terminal.In practical wireless networks, no matter for a base station, a relaynode or a user node, there is always imperfect instantaneous channel stateinformation(CSI) due to estimation error, network delay or limitedfeedback, etc. This means that there exists errors between the real CSI andthe known CSI. Conventional designs which are based on perfect CSI willdegrade when encountering the imperfect CSI. Therefore, this paperdesigns and proposes a beamforming method in a MIMO relay network ormultiuser selection network with imperfect CSI, and then optimize thisbeamforming to upgrade its performance when imperfect CSI occurs. Theoptimization method aims to optimize the regularized factor in thebeamforming matrix to maximize the sum rate of the system.With comparison from the simulation and the theoretical analysis, it isobserved that the proposed method enhances the sum rate, and hasrobustness under imperfect CSI. The experiments also shows thesimulation results are consistent with the theoretical data. Finally, theobtained robust beamforming provides ideas, methods and conclusion for reference for the further research in the similar model with imperfect CSI.The CSI error is modeled as Gaussian distributed, therefore, based onwhich, some specific model can be further studied for the future work.