Outage-optimal relay selection for energy-harvesting relays based on power splitting

Radio frequency (RF) based wireless charging and information transfer, also known as simultaneous wireless information and power transfer (SWIPT), is a new paradigm of green-energy wireless communications. In this work, we apply SWIPT to cooperative communication networks where each relay is equipped with an energy harvesting (EH) circuit to extract and convert the RF signal into useable energy for transmission. Since EH and baseband circuits can not process the same signal, time-sharing (TS) and power-splitting (PS) receiver structures are commonly considered but their optimal settings have not been explored. Focusing on the PS receiver structure, this work starts from identifying the optimal PS factor that minimizes the outage probability. Since the optimization of the PS factor requires the knowledge of the instantaneous channel gains that need to be updated frequently, we resort to an approximation analysis, which optimizes the PS factor using the first-order statistics of the fading channels. Simulation results are presented to validate the approximated optimal PS factor. Finally, three relay selection schemes employing the optimal PS factor are evaluated and compared in terms of their average throughput.