Border Collie optimization algorithm‐based node clustering technique in vehicular ad hoc networks

Abstract

AbstractIn vehicular ad hoc networks (VANETs), the nodes are vehicles and these vehicles communicate with one another on the road. The vehicular nodes on the road move continuously and it introduces dynamic changes in the topology of the network. When the node density in the vehicular network is high, then it is still more challenging as these situations introduce different challenges for network scalability and hurdles the process of determining an optimal route in the vehicular network. At this juncture, clustering protocols are often utilized for resolving the issue of network scalability and the factors that hinders process of identifying optimal routes in vehicular network. In this article, Border Collie optimization algorithm‐based node clustering technique (BCOA‐NCT) is proposed for achieving optimal cluster head selection that minimizes network overhead in uncertain node density situations. This BCOA‐NCT is proposed for facilitating equal concentration on both exploitation and exploration during the process of clustering. It utilized the strategy of eyeing for preventing the algorithm from being stuck into local optima during the cluster head selection process. The simulation experiments of the proposed BCOA‐NCT and the benchmarked schemes confirmed better throughput of 6.19%, minimized end‐to‐end delay of 7.24%, reduced mean energy consumptions of 6.89%, and reduction in the number of beacon messages utilized by 8.26% under the impact of different density of vehicular node. The outcome of the results indicated that the proposed BCOA‐NCT on par with the baseline approaches is capable enough in handling the issue of scalability and better estimation of optimal routes under the increase of node density in the vehicular network.