A Hybrid Post-Quantum Cryptography and Machine Learning Framework for Intrusion Detection in VANETs

Abstract

Vehicular Ad Hoc Networks (VANETs) are a fundamental technology for intelligent transportation systems, enabling real-time communication between vehicles, roadside infrastructure, and cloud-based services. However, the increasing connectivity of vehicles introduces significant cybersecurity challenges, including replay attacks, Sybil attacks, false message injection, and denial-of-service attacks. Existing VANET security mechanisms primarily rely on conventional cryptographic algorithms such as RSA and ECC, which are vulnerable to future quantum computing threats. Furthermore, traditional intrusion detection systems lack the intelligence and adaptability required to detect evolving cyber threats in highly dynamic vehicular environments. This paper proposes a Hybrid Post-Quantum Cryptography and Machine Learning Framework for Intrusion Detection in VANETs (HPQC-ML-VANET). The proposed framework integrates post-quantum cryptographic mechanisms based on lattice-based algorithms with a machine learning-driven intrusion detection system. The cryptographic layer provides quantum-resistant authentication and secure key exchange, while the machine learning layer performs behavioural analysis and real-time anomaly detection. The proposed approach combines CRYSTALS-Kyber for secure key encapsulation and CRYSTALS-Di lithium for digital authentication with a hybrid machine learning model based on feature extraction, dimensionality reduction, and ensemble classification. The framework is evaluated using simulated VANET communication environments incorporating realistic vehicle mobility and cyberattack scenarios. Performance evaluation considers detection accuracy, false positive rate, communication overhead, and detection latency. The proposed framework aims to provide a scalable, privacy-preserving, and quantum-resistant security solution for future autonomous and connected vehicles.