Sum Power Minimization of Greener Communications with Energy Harvesting

Energy harvesting (EH) plays an important role in greener wireless communication systems. Causality of EH increases challenge. Guaranteeing the throughput as a quality of service (QoS), we aim at using the least power, e.g. for broadband satellite communications. Solving this target problem requires exactness and efficiency. The proposed generalized water-filling approach computes the solution to the sum power minimization problem while meet EH and QoS constraints with K epochs. The proposed algorithm with the assumption of predictable channel power gains and incoming harvested energy, possesses distinguished features: exact numerical values of the solution, i.e., the exact solution, to the target problem; and a low degree polynomial complexity, not beyond O(K
^2.2
), which is lower than a cubic polynomial complexity in K. Since EH problems are new emergence, the conventional water fillings cannot solve the proposed problem, under the merit of exactness and less complexity. The two advantages, exactness and the efficiency, are very suitable for a real-time optimal power allocation. Really, the proposed algorithm is simple, but the reason of its optimality is profound and then omitted here. Optimality of the proposed algorithm is guaranteed. Numerical results illustrate the steps and demonstrate the efficiency of the proposed algorithm. To the best of the authors' knowledge, there is no existing reference to provide such a solution under the merit of the mentioned two advantages, including the most popular interior point method.