Neural Network-Based Intelligent Routing for Secure VANET Communication

Abstract

Transportation Systems (ITS) enable seamless communication between vehicles and roadside infrastructure. This connectivity significantly enhances road safety, traffic efficiency, and overall driving enjoyment for users. However, router protocols in VANETs encounter substantial challenges due to the high mobility of vehicles and the rapid changes in network topologies. Traditional routing methods often suffer from delays and packet loss as a result of these dynamic conditions. To address these issues, we propose a novel algorithm that leverages machine learning techniques, specifically utilizing neural networks for intelligent routing in VANETs. This innovative approach dynamically optimizes routing decisions while also enhancing communication security. By effectively detecting and mitigating potential attacks, our algorithm improves routing efficiency, reduces communication delays, and strengthens data security. Simulation results indicate that our proposed system outperforms existing routing protocols, leading to improved network performance and a significant reduction in end-to-end delay, particularly in challenging scenarios such as black hole attacks.