Enhancing Intra-Vehicle Network Security using Intelligent Data-Driven Models

Electric vehicles experience critical operational challenges because they depend greatly on vehicle-based communications and inter-vehicle network communications. The research examines electric vehicle cyber security threats followed by a recommendation of an intelligent secure framework which defends against unauthorized access. This model adopts an upgraded support vector machine anomaly detection system which uses controller area network bus protocol. The model requires additional innovative developments to enhance performance for fast malicious attack detection and mitigation. It employs a novel optimization algorithm, but this method uses social spider optimization techniques for operation. This approach develops a technique that enhances the model's offline training process. The two-stage modification the system incorporates a dual-stage system design to enhance optimization performance by improving search capabilities and minimizing premature convergence risks. The model shows outstanding performance alongside reliable security and high accuracy against denial-of-service attacks in electric vehicles through results obtained from real-world datasets.