Optimal power control for energy harvesting cognitive radio networks

Cognitive radio (CR) can be combined with energy harvesting to lift the spectrum efficiency and make use of green energy. The allocated power for the secondary user (SU) needs to have peak power constraints to restrict the interference with the primary user (PU). On the other side, the energy harvesting property of the nodes leads to the causality feature when allocating the harvested energy. In this paper, we apply our recently proposed geometric water-filling with peak power constraints (GWFPP) and recursion machinery to solve the target throughput maximization problem. The proposed algorithm is precisely defined. It provides the exact optimal solution via efficient finite computation. Optimality of the proposed algorithm is strictly proved. Numerical results are presented to illustrate steps and effectiveness of the proposed algorithm, and the exact optimal solutions to the problem. Significant throughput gain can be observed over the well-known primal-dual interior point method (PD-IPM), which only guarantee to generate an ∈ solution to the problem.