Analysis of Probabilistic Forwarding in Opportunistic Networks Using the Percolation Theory

Abstract

Opportunistic networks are mobile networks that rely on the store-carry-and-forward paradigm, using contacts between nodes to opportunistically transfer data. For this reason, traditional routing mechanisms are no longer suitable. To increase the success of successful message delivery, different probability-based techniques were previously studies by various authors. Here we address the question of how much of the forwarding probability of an ON has to be increased, in order for the network to achieve a given desired hit rate. We propose an approach based on percolation theory, which explains the influence of forwarding probability in a network's performance, and we try to prove that such a phenomenon is indeed present in ONs. We demonstrate, through extensive experiments, that the transition phase can be indeed observed in ONs when the forwarding probability is varied from 0 to 0.1. After the transition phase, little benefit is obtained in terms of reachability (exponential relationship) when the forwarding probability is increased. In contrast, the delivery cost increases much faster than the reachability after the transition phase. Consequently, increasing the forwarding probability only impacts on metrics like the delivery cost and latency since high reachability can be assumed in opportunistic networks.