MANET Network Efficiency Optimization Using the PSO-FFNN Node Clustering Method

Abstract

The performance of a mobile ad hoc network (MANT) is impacted by both the traffic volume and the speed of the individual nodes. Traditional routing strategies have been devised in order to increase performance measures such as throughput and latency; nevertheless, there is still opportunity for improvement, particularly when routing is directly tied to the amount of energy that a node consumes. In this investigation, the process of routing is carried out by utilizing the clustering and superclustering techniques. There are four cluster heads that connect the base station node to hosting nodes that have 100, 200, and 500 numbers respectively. A technique known as far-near base is utilized in order to accomplish the goal of establishing communication between the nodes. The improved clustering method makes use of a Feed Forward Neural Network that is based on Particle Swarm Optimization (PSO) (FFNN). According to the results that have been presented, the performance of the network has been enhanced; the throughput in PSO-FFNN has grown by 88.839 percent, 92.73 percent, and 83.224 percent, respectively, for 100, 200, and 500 nodes.