Online Joint Power Control for Two-Hop Wireless Relay Networks With Energy Harvesting

We consider a two-hop amplify-and-forward relay network with energy harvesting nodes, and design online joint power control at the source and the relay to maximize the long-term time-averaged rate over fading channels. We formulate the problem as a joint stochastic optimization problem under battery operational constraints and finite storage capacity constraints. In seeking an online solution, we transform the problem into one that enables us to leverage Lyapunov optimization to develop an online algorithm to provide the joint power control solution for the source and the relay in a fading environment. The joint power control solution is derived in closed-form and only depends on the current energy arrival at each node and fading condition over each hop, without requiring any statistical knowledge of them. Our proposed algorithm not only adapts the power based on the battery energy levels to conserves energy, but also exploits opportunistic transmission based on fading condition. Through analysis, we show that the performance gap of our proposed algorithm to the optimal power control policy is bounded. Simulation results demonstrate a significant gain of our proposed online joint power control algorithm over other alternative methods, including pernode separate power control and heuristic joint power control methods.