Optimal power allocation for AF full-duplex relay in cognitive radio networks

This paper investigates the power allocation between the cognitive transmitter (CT) and the cognitive relay (CR) in a cognitive Amplify-and-Forward (AF) full-duplex relay network. To maximize the cognitive capacity, we firstly obtain a closed form expression of the optimal power allocation with the instantaneous interference channel information. Then considering limited cooperation between the primary and the cognitive networks, we develop an efficient iterative search algorithm to obtain the optimal power allocation with the statistical rather than the instantaneous interference channel information. Furthermore, considering absolutely no cooperation between the primary and the cognitive networks, we develop a sensing-based iterative search algorithm to obtain the optimal power allocation without any interference channel information. Simulation results show the capacity of the cognitive AF full-duplex relay network may achieve up to 50% improvement compared with the cognitive AF half-duplex relay network on the condition that the mean value of the self-interference channel gain is lower than -25 dB.