EFFICIENT ALGORITHM DESIGN FOR NETWORK ATTACK IDENTIFICATION TO ENHANCE V2X COMMUNICATION SECURITY IN INTERNET OF VEHICLES

Abstract

The fast growth of Internet of Vehicles (IoV) and Vehicle-to-Everything (V2X) communication offers safer and more efficient transportation systems by letting vehicles, infrastructure, and cloud systems share important data in real time. But because V2X networks are open and spread out, they can be hit by complex cyberattacks like spoofing, denial-of-service (DoS), data tampering, and Sybil attacks, which can seriously hurt safety and trust. So, one of the most important parts of making IoV settings safe is making sure that attacks are found quickly and correctly. This article shows a useful set of algorithms for finding network attacks that can be used to make V2X connection safer. The suggested method combines lightweight feature selection with graph-based anomaly detection and hybrid machine learning classifiers. This makes sure that the method is accurate at finding things that aren't normal and doesn't require a lot of extra processing power. It works well in real-time vehicle settings. The algorithm automatically tells the difference between bad and normal communication behaviours while taking into account latency limits. It does this by using temporal traffic patterns, trust scores, and adaptive thresholding. The proposed Hybrid approach gets 95.8% accuracy, 94.7% precision, 95.2% recall, 95.0% F1-score, and 96.3% AUC in experiments using benchmark IoV security datasets. This is 6–9% better than traditional detection methods. The algorithm scales well and is stable in high-motion and changing network densities. The framework integrates OBUs and RSUs without impacting gearbox performance, making it beneficial. This project aims to safeguard V2X communication by developing a rapid, flexible, and reliable attack identification system. This strengthens IoV ecosystems against new cyberattacks.