Dynamic Learning Network-Distributed Exploration: Advancements in Hybrid Swarm Intelligence Algorithm for Congestion Control in Vehicular Ad-Hoc Networks

At the Nexus of Intelligent Transportation Systems (ITS), Vehicular Ad-hoc Networks (VANETs) have become a quickly developing field, highlighting the necessity of a resilient and reliable VANET architecture to support increasing vehicle densities. This study has proposed a new technique, namely a Hybrid Swarm Intelligence Algorithm (HSIA), that mixes distributed exploration approaches rooted in hybrid swarm intelligence paradigms and dynamic learning mechanisms. Our proposed approach, which builds on the ideas of Artificial Bee Colony (ABC) and Ant Colony Optimization (ACO), combines network-distributed communication, dynamic learning rates, and reinforcement learning approaches to improve algorithm performance and adaptability. We provide novel equations for adaptive pheromone updates with Dynamic Learning, Collective Exploration with Network-Distributed Pheromone Update and Adaptive Exploration-Exploitation Trade-off with Reinforcement Learning. Our experimental results on VANETS show that our method is more controlling and versatile than conventional ACO or ABC algorithms when striving for quicker convergence rates and higher-quality results. MATLAB simulations are used for the experimental validation of the HSIA technique, which shows improved performance over conventional ACO and ABC. Comparing HSIA to ACO, the packet delivery ratio and throughput increased significantly to 1.09 percent and 1.48 percent, respectively. Compared to ABC, HSIA showed an incredible increase in its packet delivery ratio and throughput of 15.94% and 9.87%, respectively. HSIA experienced a 17.25% lower end-to-end delay compared to ABC. On the other hand, HSIA’s end-to-end delay was 5.59% lower than that of ACO. Critical performance metrics showing this improvement include packet delivery ratio, throughput, and end-to-end delay.