A Model for Self-Adaptive Routing Optimization in Mobile Ad-Hoc Network

Abstract

This study designs, simulates and assesses the performance of a Self-Adaptive Partitioned Particle Swarm Optimization (SAP-PSO) routing model in a MANET. The model automatically groups nodes into partitions and obtains the local best for each partition. The local best for each partition communicates with each other to form the global best. The model was simulated and benchmarked with the Traditional PSO (T-PSO) and the Ant Colony Optimization (ACO) using global best and computational time as performance metrics. Simulation results showed that the T-PSO and SAP-PSO does not have any significant difference in performance when there is no intermediate node on the network. The T-PSO outperformed both ACO and SAP-PSO models when intermediate nodes on the network were few. When a large number of intermediate nodes are present on the network, the proposed SAP-PSO performed better than PSO and ACO. This makes SAP-PSO a better routing optimization when large numbers of intermediate nodes are on the network and the search space is complex.