Outage performance of relay selection with spatially random relays using RF energy harvesting

To enable green wireless networks, one appealing approach is to deploy energy harvesting (EH) relays to assist the source transmission. Unlike conventional relays relying on fixed power supplies, EH relays make use of the energy collected from the RF radiation of the source node, and thus, they do not introduce extra energy cost to the network. This paper presents an analytical study to assess the efficacy of EH relays when the one with the maximal end-to-end signal-to-noise ratio is selected to perform data relaying while others perform EH. Because the action (either harvesting energy or forwarding data) of one EH relay affects those of others, exact performance analysis is not tractable. Additionally, relay density and positions may be random, which further complicates the analysis. Our analysis is conducted based on the hypothesis that each EH relay has an equal chance to be selected. This hypothesis allows for analytical tractability and is of importance to EH relays because otherwise some may drain their batteries fast. We identify the conditions under which the aforementioned hypothesis is valid. Our analysis also considers two variants of amplify-and-forward relays with and without using channel state information. Numerical results are presented to validate the analysis accuracy along with extensive discussions on the impact of numerous system parameters.