A novel interference-aware, optimal gain adaptation policy for a non-regenerative, underlay cognitive radio relay

A relay that operates in an underlay cognitive radio (CR) system is subject to tight constraints on the interference its transmissions cause to the primary receiver (P_Rx). While several appealing non-regenerative amplify-and-forward (AF) relaying schemes have been proposed for underlay CR, they are simple, ad hoc adaptations of conventional AF relaying schemes. We present a novel and optimal relay gain adaptation policy (ORGAP) in which the interference-aware relay optimally adapts its gain as a function of its local channel gains. It minimizes the symbol error probability (SEP) at the secondary receiver subject to a constraint on the average interference caused to the P_Rx. We also analyze the SEP of MPSK of ORGAP, which serves as a fundamental theoretical benchmark for the gains achievable by AF relaying in CR systems. Extensive numerical results quantify the extent to which ORGAP outperforms several known conventional AF relaying schemes. We also present a near-optimal, simpler relay gain adaptation policy (SRGAP) that is easier to implement.