A Swarm Intelligent–Based Cluster Optimization in Vehicular Ad Hoc Networks for ITS

Abstract

The Internet of Things (IoT) has transformed vehicular ad hoc networks (VANETs), leading to the Internet of Vehicles (IOV). VANETs are wireless networks without fixed infrastructure, designed to improve traffic safety in real time, supporting intelligent transportation systems (ITS). Due to their unpredictable nature, VANETs face major challenges like frequent link failures, scalability, reliability, network layout issues, quality of service (QoS), and security, all of which are complex and difficult to solve (NP-hard problems). Traditional protocols are unsuitable for VANETs due to their unique properties. To accomplish the optimal number of clusters and achieve stability in VANETs within a dynamic environment, we propose a swarm-based metaheuristic algorithm called the rat swarm optimization (RSO) algorithm. The RSO algorithm employs a clustering technique to optimize the network performance and ensure efficient communication in VANETs. The RSO algorithm optimizes load based on node transmission range (Tx range) through effective resource utilization and coordination. RSO organizes the unstructured network into cluster structures and generates near-optimal clusters and CHs to reduce network randomness and maintain stability with lower communication costs. By keeping the number of clusters at an optimal level, the RSO algorithm enhances cluster lifetime and overall network performance. To assess the effectiveness and efficiency of the RSO algorithm, numerous experiments are performed by using various grid sizes, Tx ranges, and nodes in the network. The generated results demonstrate that the RSO algorithm stimulates 50.96%, 33.15%, 88.73%, and 96.70% optimal number of clusters when contrasted with the clustering algorithm–based on ant colony optimization (CACONET), moth flame clustering algorithm for IoV (MFCA-IoV), the whale optimization algorithm for clustering in vehicular ad hoc networks (WOACNET), and grasshoppers' optimization-based node clustering technique for VANETs (GOA) when the Tx range and nodes are taken into consideration. But, when the grid size is considered, the RSO generates 32.31%, 15.23%, 47.04%, and 58.33% optimal number of clusters when compared to cutting-edge algorithms. Hence, the quantitative results and the statistical representation show the proposed RSO algorithm's effectiveness over cutting-edge algorithms under the unpredictable nature of VANETs.