Semi-blind iterative relay selection schemes for amplify-and-forward cooperative relay network

Optimal power allocation (OPA) schemes for amplify-and-forward (AAF) relay networks have been widely investigated in recent years, usually under the assumption of readily known channel state information (CSI). This paper proposes a semi-blind iterative relay node selection (SBIRS) scheme to carry out the OPA procedure when the CSI are entirely unknown at the destination. We first develop a semi-blindly weighted multiple-eigenvector (SBWMEV) solution for optimally allocating power among all the relay nodes in the network. Aided by the intrinsic salient feature of the SBWMEV eigenvector structure, we further propose an iterative relay selection algorithm to simplify the SBWMEV solution with enhanced performance. In an iterative relay node selection manner, the proposed SBIRS scheme is capable of potentially allocating more power to the relay nodes with better channel conditions. Simulations verify that the SBIRS schemes outperform the SBWMEV scheme in terms of throughput and outage probability.