Relay Selection for the Wireless Energy Harvesting Relay Assisted Network

Relay assisted (RA) network with energy harvesting (EH) is a promising method to realize the green communication. Meanwhile, the relay selection (RS) is an effective method to exploit the performance advantage of EHRA network. However, the performance of RS depends on the perfect channel state information (CSI) deeply. And, there is an obvious performance gap between the RS strategy with perfect CSI and partial CSI. In order to improve the performance of EH-RA network with partial CSI, a RS strategy is proposed to optimize the spectrum efficiency (SE) under an EH constraint in this paper. The upper bounder (UB) of the proposed RS strategy is also analyzed under the perfect CSI assumption. Moreover, the statistical dynamic program (SDP) is investigated to cooperate with the proposed RS strategy by optimizing threshold under the partial CSI assumption. Numerical results show that the SE performance of the proposed RS strategy with SDP is much better than the greedy algorithm and approaches to the performance UB effectively.