HYBRID ADAPTIVE OPTIMIZATION ALGORITHM COMBINATION OF PARTICLE SWARM OPTIMIZATION & GREY WOLF OPTIMIZATION WITH REINFORCEMENT LEARNING (HAPSOGWO-RL) FOR EFFICIENT DATA ROUTING IN VEHICULAR AD HOC NETWORK

Vehicular Ad hoc Network (VANET) consists of a number of vehicular nodes that can independently join or leave the network at any time. Due to the high mobility of vehicles, establishing a secure and efficient communication route between two endpoints becomes a complex challenge. This paper presents a routing protocol for vehicular ad hoc networks that utilizes a hybrid adaptive optimization algorithm, combining and Adaptive Particle Swarm Optimization (APSO) and Adaptive Grey Wolf Optimization (AGWO)) techniques. In this proposed model, the entire network is divided into clusters, and the hybrid adaptive optimization algorithm is employed to determine the Cluster Head (CH) within each cluster. The hybrid AGWOPSO-RL algorithm operates iteratively to select the most suitable vehicle to serve as the cluster head, ensuring stable and energy-efficient communication. Furthermore, reinforcement learning is applied within the network to make dynamic decisions regarding vehicle cluster changes, thereby enhancing route adaptability and stability. Data transmission between the source and destination is carried out through the selected cluster heads, optimizing both communication reliability and efficiency. The proposed protocol is implemented and simulated using Network Simulator 2 (NS2), and the results are analysed in terms of multiple performance metrics, including packet delivery ratio, end-to-end delay, throughput, and routing overhead. The experimental outcomes demonstrate that the proposed Adaptive Hybrid APSO-AGWO-based routing protocol achieves better network stability, higher packet delivery, and improved overall performance compared to conventional routing approaches