INTRAVEHICLE COMMUNICATION SECURITY USING ADVANCED MACHINE LEARNING

Abstract

ABSTRACT:The high reliance of electric vehicles on both in-vehicle and between-vehicle communications can cause significant issues in the system. This paper addresses the issue of cyber attacks on electric vehicles and proposes a secure and reliable intelligent framework to prevent hackers from penetrating the vehicles. The proposed model is constructed based on an improved support vector machine model for anomaly detection, utilizing the Controller Area Network (CAN) bus protocol. To enhance the model's capabilities for fast malicious attack detection and prevention, a new optimization algorithm based on the Social Spider Optimization (SSO) algorithm is developed, which reinforces the offline training process. Additionally, a two-stage modification method is proposed to increase the search ability of the algorithm and avoid premature convergence. The simulation results on real data sets demonstrate the high performance, reliability, and security of the proposed model against denial-of-service (DoS) attacks in electric vehicles.