Optimal Recursive Power Allocation for Energy Harvesting System With Multiple Antennas

In this paper, we investigate the optimal recursive power-allocation policies with energy harvesting wireless nodes equipped with multiple antennas in a fading channel. This optimization problem includes several complex matrices as optimization variables. As a difference, existing optimization theory and methods have been designed to solve these problems over real space. Naturally, the optimization variables have been assumed points in the real space. We proposed a transform approach and designed the algorithms for solving the throughput maximization problem and transmission completion time minimization problem for a multiple-input-multiple-output (MIMO) system. The algorithms were further extended to solve the throughput maximization problem of a hybrid system with both harvesting energy and grid power. Numerical results illustrated the algorithm steps and significant efficiency of the proposed algorithms. To the best of the authors´ knowledge, there are no existing algorithms reported in the open literature to obtain exact solutions to the proposed problems. Significant features of the proposed algorithms are that 1) they provide the exact optimal solutions via efficient finite computation and that 2) optimality of the proposed algorithms is strictly proven.