Routing approaches and performance evaluation in delay tolerant networks

Abstract

In this paper the problem of routing in Delay Tolerant Networks (DTNs) is considered. In these networks, where nodes are sparse and the links between them are intermittent, there is no guarantee that a path between source and destination nodes exists at any time, rendering traditional routing protocols unsuitable to deliver messages between nodes. Different DTN routing protocols for stochastic or dynamic networks, where the time-evolving topology is random and not known, are examined. These protocols depend on decisions regarding where and when to forward messages. Performance metrics such as message delivery ratio, average latency, buffer occupancy and average hop count are analyzed in order to compare Direct Delivery, First Contact, epidemic routing and the VACCINE recovery mechanism, Spray and Wait protocols, PROPHET and MaxProp. Simulations show that MaxProp and epidemic protocol combined with VACCINE are able to deliver more messages than the other protocols, ensuring also a lower delivery latency and a lower percentage of buffer occupancy.