A General Connectivity Model for Non-Linear SWIPT Systems With Spatially Randomly Distributed Relays

In the context of fully distributed systems, we consider a scenario densely populated with wireless nodes which are equipped with simultaneous wireless information and power transfer (SWIPT) capabilities and can act as relays between a source and a destination. In this scenario, the probability of finding a node able to provide a useful contribution to the quality of the link tends to be high and can be exploited adopting a suitable relay selection scheme. Assuming nodes distributed following a Poisson point process and communicating only with source and destination to ensure the scalability of the system, we consider two well-known relay selection schemes, which are typically used as performance benchmarks, and derive an analytical framework to investigate the performance of the source-destination link in terms of outage probability. In our analysis we include, through the use of a simple but effective model, the non-linear effects of the energy harvesting circuitry of the devices and introduce a simple signaling mechanism between source, SWIPT nodes and destination to identify the most suitable relay among the various SWIPT devices. The impact of multiple access during the phase of relay selection is also modeled and investigated.