A Hybrid Optimization Framework Using Aquila and Attention Residual Net for Intrusion Detection in Vanets

This research presents a novel hybrid intrusion detection framework that integrates the Aquila Optimization Algorithm (AOA) for the shortest path selection and the Firefly Algorithm (FFA) for efficient anomaly detection in VANETs. AOA is employed to optimize routing by dynamically selecting the shortest and most energy-efficient path, reducing latency and improving network performance. Simultaneously, FFA enhances intrusion detection by leveraging swarm intelligence to detect malicious nodes based on network behavior patterns. The hybrid approach addresses the Nondeterministic Polynomial (NP)-hard nature of clustering and routing, ensuring an optimal balance between energy efficiency and detection accuracy. The proposed evaluated using key performance metrics. Simulations conducted in NS-2 and Python demonstrate that the method improves accuracy of $\mathbf{9 2 \%}$ and also enhances the packet delivery ratio and throughput and also lowers the latency and packet drop.