IMAN: An Intelligent MANET routing system

Abstract

Currently, in any MANET, one protocol routes packets to their destination regardless of the network context. This ‘one size fits all’ approach is far from optimum. Therefore, in this paper, we introduce an Intelligent MANet routing protocol system (IMAN), which employs Genetic Algorithms to select the most optimum protocol based on the network context. Comprehensive performance models of three MANET routing protocols (AODV, DSR and OLSR) were built based on the Neuro-Fuzzy Inference System. Representative MANET scenarios, with different number of nodes and mobility schemes, were tested by means of simulations. Findings indicated considerable reduction in packets delay and data load when IMAN was utilized.