Stochastic Performance Evaluation of Routing Strategies in Opportunistic Networks

The opportunistic networks represent a new communication paradigm born from mobile ad hoc networks and delay-tolerant networks. This paradigm is rooted in the application level and builds the network based on the simple concept of connecting nodes to one-hop at a time. This means that nodes in the network are not aware of the destination path, i.e. they just pass the message to the next node. Understanding their behavior will help us to identify the best way to propagate messages depending on the situation. Most studies that try to understand these networks behavior use simulations or empirical tests. Although these approaches are useful, they involve an important e ort and do not provide enough exibility to explore the network behavior in an easy and fast way. This paper presents an analytical model of an opportunistic network as a way to overcome this limitation. Protocol designers and developers of communication infrastructures can take advantage of this model to determine the best way to disseminate messages in opportunistic networks, according to particular communication conditions. Two routing strategies for these networks have been formalized in terms of the number of copies and hops allowed for a message. The performance of these routing algorithms was evaluated considering three variables: mean time to arrival of a message, expected number of message copies at the delivery time, and energy consumed in the message transmission. Thus, we show the usability and usefulness of the proposed analytical model.