Skip to Main Content
This paper explores the balance between cooperation through relay nodes and aggregated interference generation in large decentralized wireless networks using decode-and-forward. The source nodes in the network are modeled using a marked Poisson process. We consider the case in which only a single randomly located relay is added to one source in the network and study the outage probability gains obtained. Then, using a simple model, we study the case in which all sources can potentially use their nearest neighbor from the set of inactive nodes as relays, leading to a mixed transmission scheme in which some users employ decode-and-forward and others employ direct transmission. The optimal relay activation probability for the second case is found, observing that in the small outage probability regime it exhibits a binary behavior, being zero or one. Comparing both scenarios we conclude that activating more relays rapidly reduces the gains observed when only one source can use a relay. We derive closed-form approximations to the upper bounds on the error probability, averaging over all node positions and fading gains realizations, to support our claims.