Real Time Anomaly Detection and Intrusion Detection for Safeguarding Intra-Vehicle Communication Powered by AI

This study addresses cyber-attacks in Electric Vehicles (EVs) and proposes an intelligent, secure framework to protect both in-vehicle and vehicle-to-vehicle communication systems. The proposed model uses an improved support vector machine (SVM) for anomaly and intrusion detection based on the Controller Area Network (CAN) protocol a critical component in vehicle communication. To further enhance detection speed and accuracy a new optimization algorithm the Social Spider Optimization (SSO) is introduced for reinforcing the offline training process. Simulation results on real-world datasets demonstrate the model's high performance, reliability and ability to defend against denial-of-service (DoS) attacks in EVs.