A geographic routing algorithm based on Cat Swarm Optimization for vehicular ad-hoc networks

Geographic routing protocols require only local information for routing decisions and thus is considered very efficient in large-scale wireless networks such as Vehicular Ad-hoc Networks (VANETs). In geographic routing, one of the neighbor of sender vehicle is selected as next forwarding vehicle to forward the message to the destination nodes. Various geographic routing protocols have been proposed in literature for efficient selection of next forwarding vehicle. However, these protocols show inadequate performance due to unique characteristics of VANETs such as high mobility, low bandwidth, and restricted movement. In this paper, a novel geographic routing (GR) protocol based on Cat Swarm Optimization (CSO), namely CSO-GR is proposed that aims to efficiently find optimized position of next forwarding vehicle in highly dynamic vehicular environment. A fitness function to optimize impact of various parameters on the selection of next forwarding vehicle is suggested. Performance of the proposed protocol is assessed via simulation. Significant improvement can be seen in the performance of proposed protocol with regard to packet delivery ratio and normalized routing load.