Collaborative Detection Framework Using ML for VANET Malicious Node Localization

Vehicular ad hoc networks (VANETs) play a vital role in enabling intelligent transportation systems by facilitating real-time communication among vehicles. However, the open and dynamic nature of VANETs makes them highly vulnerable to malicious nodes, which can disrupt communication, compromise safety, and degrade network performance. This paper proposes a collaborative detection framework that leverages machine learning (ML) techniques for the accurate localization of malicious nodes in VANETs. The framework integrates vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) data to enhance situational awareness and anomaly detection. By applying supervised and semi-supervised ML models, such as Random Forest and Semi-Supervised Support Vector Machines (S3VM), the system identifies behavioral patterns that indicate malicious activity. Additionally, trust evaluation and data correlation across multiple nodes improve detection accuracy and reduce false positives. Experimental simulations conducted on real-world traffic scenarios demonstrate the robustness and adaptability of the proposed approach, showing significant improvements over traditional rule-based methods. This research contributes to safer and more reliable VANET operations by introducing a scalable and intelligent solution for localizing malicious nodes.