Throughput maximization in two-hop energy harvesting communications

Two-hop energy harvesting communications are considered. The scenario consists of a source node which wants to send data to a destination node through a half-duplex amplify-and-forward relay station. The source node and the relay station harvest energy from the environment several times and use it to transmit the data. Our goal is to find the optimal power allocation that maximizes the throughput at the destination node. We show that the use of a half-duplex amplify-and-forward relay station leads to a non-convex optimization problem. Therefore, to find the optimal power allocation we propose to reformulate the problem as the difference between two concave functions (D.C. programming). Moreover, a branch-and-bound algorithm is tailored to fit the energy harvesting constraints. We show that the feasible region has to be adapted to facilitate the branching process. Additionally, we reduce the complexity in the calculation of the bounds by relaxing the problem into a convex problem with a linear objective function. Numerical results compare the performance in different energy harvesting scenarios.