An adaptive genetic fuzzy multi-path routing protocol for wireless ad-hoc networks

The inherent uncertainty in wireless mobile ad hoc networks (MANET), due to nodal mobility, unstable links, and limited resources, frequently renders routing paths unusable. Thus, recurrent route discoveries detrimentally affect network performance. The most promising solution is to use multiple redundant paths for routing. However, selecting an optimal path set is a NP hard problem. Most current multi-path routing protocols do not concentrate on the uncertainty in MANET. They choose an "optimal" multi-path set by considering only one single route selection parameter, such as the least number of intermediate hops or the maximal remaining battery power. As a result, they miss the correlations among the multiple route selection parameters. This paper proposes the genetic fuzzy multi-path routing protocol (GFMRP), which is a multi-path routing protocol based on fuzzy set theory and evolutionary computing. GFMRP naturally deals with the uncertainty in MANET and adaptively constructs a set of highly reliable paths by considering the interplays among multiple route selection parameters. GFMRP takes into account four important factors as the selection parameters; which are the energy consumption rate, queue occupancy rate, link stability, and the number of intermediate nodes. The performance of GFMRP is evaluated in terms of packet delivery ratio, average end-to-end delay, and the frequency of route rediscovery in ns2 context. Simulation results demonstrate that GFMRP is well suited to the ad hoc environment and outperforms DSR, SMR and SBMR.