Cooperative Relaying with Imperfect Channel State Information

We consider relay cooperation with imperfect channel state information (CSI) in the downlink of wireless networks. In particular, we consider a two-phase transmission where in the first phase the base station broadcasts information to the relays; the relays decode the data fully or partially depending on the transmission rate and the quality of their corresponding communication links. During the second phase, the relays cooperate by jointly beamforming information to multiple users given that channel mean and covariance are available at the transmitter side. The goal is to optimize the total network throughput (taking into account both transmission phases) by proper choice of the transmission rates, cooperation architecture and beamforming transmit vectors from the relays. The key contribution of this paper lies in the consideration of the impact of CSI imperfections in such a system. We first formulate the problem of finding the optimum throughput, which is not amenable to analytical solution. We therefore derive a suboptimum adaptive beamforming strategy that maximizes a derived upper bound on the average system throughput. Even though the relays have imperfect CSI, it is shown that relay cooperation can significantly improve the overall system throughput.