Resource allocation for relay-assisted cognitive radio network with multiple channels

Resource allocation for relay assisted cognitive radio network was investigated in this paper. Interference temperature limits of corresponding primary users were considered. An optimal joint channel and power allocation for a centralized scenario was presented in this work. Generalized expectation maximization (GEM) theory based optimization algorithm was presented to settle the non-convex power allocation problem. It broke the non-convex problem into two convex sub-problems and obtained the solutions iteratively. Numerical results are given to verify the availability of the power allocation algorithm. It shows that the end to end throughput can be enhanced by joint channel and power allocation. It also shows that channel allocation is more efficient than power allocation. An interesting phenomenon is found that the most efficient resource allocation scheme varies with interference temperature limits of primary users.