AI-based Sensor Attack Detection and Classification for Autonomous Vehicles in 6G-V2X Environment

Document Type


Publication Title

IEEE Transactions on Vehicular Technology


Autonomous Vehicles (AVs) mainly rely on sensor data and are anticipated to transform the transportation sector. The abnormal sensor readings generated by malicious cyberattacks or defective vehicle sensors can result in deadly crashes. This paper proposes a Sensor Attack Detection and Classification (SADC) framework in a 6G-V2X environment to examine the cybersecurity concern for AVs against sensor attacks. The SADC framework employs GPS and LiDAR sensor attack detectors and the Pattern-based Attack Classification (PAC) algorithm. It combats new-age cyberattacks and provides an accurate sensor attack detection and classification mechanism in AVs. A protocol-based attack detection scheme in SADC is developed to identify the abnormal source sensor based on the detector's results. The PAC algorithm classifies malicious sensors by analyzing different strategies: instant, constant, bias, and gradual drift. The results show that the SADC framework has a 0.98% higher accuracy than the existing counterparts in detecting attacks and classifying them efficiently.



Publication Date



6G mobile communication, 6G-V2X, Autonomous Vehicles, Behavioral sciences, Computer crime, Global Positioning System, GPS and LiDAR Sensor Attack Detectors, Laser radar, PAC, Robot sensing systems, Security, Sensor Attack Detection

This document is currently not available here.