Geometric Shapes-based PSO Approach for Routing in Vehicular Networks

In communication between vehicles, many mobile nodes are linked from the source to the final destination for message transfer purpose when the vehicles are constantly moving and the topology is wide. There are scenarios in which a node wants to multicast some type of alert or specific message to nearby neighbor nodes. In this paper, an efficient routing protocol called geometric shapes-based PSO approach for routing in vehicular networks (named GPSV) is proposed, which exploits the particle swarm optimization (PSO) technique and the angle between the nodes to determine a suitable route to transfer the alerts or specific messages from a source node to a group of destinations. In the proposed method, each route request packet is considered as a swarm message which has an inertia, velocity, global best experience, personal best experience, and current position coordinates. By using the PSO technique as backbone of angular routing, the swarm determines the next best hop of a node. Simulations are conducted to evaluate the performance of the proposed GPSV protocol, showing that it can take any possible shape. We have used circular, lane, and sine waves to test the swarm formations. In addition, GPSV outperforms the flooding method in terms of packet delivery ratio, average delay, throughput, and loss packet percentage.