Probabilistic Routing Based on History of Messages in Delay Tolerant Networks

Abstract

Unexpected disconnections, long transmission latencies and network heterogeneity that DTN is meant to support make the routing problem in DTN very complicated. In DTN, probabilistic based routing protocols make use of nodes' mobility history to gauge delivery likelihood of nodes. However, the delivery likelihood of the nodes is not the only factor that affects the message delivery likelihood. In this paper, a detailed analysis of routing in DTN reveals the forwarding decisions that could go wrong due to the store-and-forward nature of DTN. Based on the analysis, we propose a history of messages concept which Probabilistic Routing Protocol using History of Encounters and Transitivity (PRoPHET) can utilize to improve the chances of message delivery. Our simulations show that using the messages' history improves the message delivery performance of PRoPHET and is comparable to MaxProp.