Choosing the objective of optimal routing protocols in Delay Tolerant networks

Abstract

Networks in which nodes are sparsely distributed and, therefore, are disconnected for long periods of time, are termed Delay Tolerant networks (DTN). The intermittent connection, together, with the limited resources of mobile nodes, mainly power and memory, created a challenging environment for data networking in DTN. Routing protocols developed for DTN tend to discover and select the minimum end-to-end delay paths to destinations assuming that these paths provide the highest delivery rate. To achieve this goal, they spread many copies of the same packet, ignoring the limitedness of storage space and power sources. In this paper, we study this problem by building a mathematical model for optimal routing in DTN. We compare the results of implementing three objectives for this model: minimizing the end-to-end delay, minimizing the end-to-end number of hops, and maximizing the delivered messages. We study and analyze the impact of varying the buffer space, the traffic load and the packets time-to-live (TTL) on the three objectives. Results show that minimizing the number of hops provides higher delivery ratio than minimizing the delay, which contradicts the previous assumption. In addition, minimizing the number of hops significantly reduces the number of transmissions which results in saving energy.