Simple ant routing algorithm strategies for a (Multipurpose) MANET model

Abstract

A Mobile Ad-hoc Network has limited and scarce resources and thus routing protocols in such environments must be kept as simple as possible. The Simple Ant Routing Algorithm (SARA) offers a low overhead solution, by optimizing the routing process. Three complementary strategies were used in our approach: during the route discovery we have used a new broadcast mechanism, called the Controlled Neighbor Broadcast (CNB), in which each node broadcasts a control message (FANT) to its neighbors, but only one of them broadcast this message again. During the route maintenance phase, we further reduce the overhead, by only using data packets to refresh the paths of active sessions. Finally, the route repair phase is also enhanced, by using a deep search procedure as a way of restricting the number of nodes used to recover a route. Thus, instead of discovering a new path from the source to the destination, we start by trying the discovery of a new path between the two end-nodes of the broken link. A broadest search is only executed when the deeper one fails to succeed. We simulated our proposal and we tuned it to the optimal performance. We also compared it with the classical approach of AODV and other biological routing approaches. The results achieved show that SARA offers the smallest overhead of all the protocols under evaluation and presents an overhead reduction of almost 25% of the value achieved by the other proposals. SARA also presents the best goodput, specially for TCP traffic, but it needs more time to discover the routes.