AIDAS: AI-Enhanced Intrusion Detection and Authentication for Autonomous Vehicles

Autonomous Vehicles (AVs) represent a transformative advancement in modern transportation systems, offering significant improvements in operational efficiency and user experience. However, their widespread implementation faces critical security challenges, particularly regarding secure remote management during system failures or cyber-attacks. These vulnerabilities potentially compromise system integrity and undermine public confidence in autonomous technologies. We introduce a novel Internet of Autonomous Vehicles (IoAV) architecture integrating an AI-driven intrusion detection system with a Chaotic Map-Based Authenticated Key Agreement protocol to address these security concerns. This integration dynamically mitigates evolving security threats through adaptive system responses. Our framework incorporates Physical Unclonable Function (PUF) technology to generate cryptographically secure private keys, establishing robust communication channels between users, Charging Stations (CS), and AVs coordinated by an Electric Service Provider (ESP). Rigorous evaluation using the Real-or-Random (ROR) model demonstrates the protocol’s resilience against diverse attack vectors, including man-in-the-middle, replay, and adversarial attacks. Experimental validation confirms the framework’s effectiveness (97.8% detection accuracy, AUC-ROC: 0.976), computational efficiency (31.25% reduction in overhead, 4.2ms inference latency), and operational resilience (99.3% authentication integrity under 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> requests/second DDoS simulation). The protocol achieves 51.38% reduced communication overhead compared to existing solutions, establishing our framework as demonstrably superior for IoAV security implementation within resource-constrained autonomous transportation infrastructures.