Genetic algorithm and probability based routing protocol for Opportunistic Networks

Abstract

The random nature of network topology is a major challenge while developing new routing protocol for Opportunistic Networks (OppNets). Additionally, other factors like intermittent connections, limited bandwidth etc. further restrict the performance of existing routing protocols. In this paper, an advanced context aware routing protocol called GAP is proposed. The proposed protocol efficiently combines the benefits of Genetic Algorithm and Probabilistic Routing to route the message from the source to destination. The protocol uses the Genetic algorithm to predict the path a message would take if it is transferred to the neighbouring node. A Fitness function is defined to evaluate the efficiency of this predicted path. The message is transferred to the neighbouring node only if the fitness value of the predicted path is greater than a threshold value, which is calculated by implementing the concepts of probabilistic routing. Simulation results show that the GAP routing protocol outperforms Prophet, Spray and Wait and GAER protocol in terms of message delivery ratio, overhead ratio and average latency.