Opportunistic amplify-and-forward relay selection with outdated channel state information

Opportunistic relay selection(ORS) is a promising approach for cooperative communications owing to its low complexity by activating only the best relay. However, in mobile scenarios where relays subject to relatively fast varying channels, the channel state information(CSI) employed in relay selection and in data transmission may be different, i.e: CSI is outdated, which may lead to wrong relay selections, and hence deteriorate the system performance evidently. Different from most literatures, which make relay selection with maximum outdated channel gain(MCG), this paper focuses on the design of relay selection strategy in amplify-and-forward(AF) cooperative networks with outdated CSI and statistical channel information for the purpose of minimizing average symbol error probability(SEP), as well as their asymptotic SEP performance analysis. Conditioned on the outdated CSI and statistical channel information, the distribution of the channel gain can be exploited. Based on this, the asymptotic SEP behavior of the system at high SNR values with each relay forwarding can be derived. Based on these asymptotic results, our selection criterion can be described as selecting the relay with the asymptotic optimum average SEP. Simulation results show that the proposed strategy achieves better performance than MCG at medium and high SNR values.