Abstract The rapid transformation of modern vehicles into interconnected Cyber-Physical Systems has significantly expan‎ded the attack surface of in-vehicle communication networks. The Controller Area Network, which serves as the core communication protocol in many vehicles, lacks inherent security mechanisms such as encryption an‎d authentication, making it vulnerable to attacks including Denial-of-Service, spoofing, fuzzy injection, an‎d replay attacks. Traditional security approaches, particularly signature-based methods an‎d cryptographic solutions, face practical limitations in automotive environments due to strict real-time requirements an‎d resource constraints of Electronic Control Units. This study presents a comprehensive review an‎d comparative analysis of Machine Learning an‎d Deep Learning-based intrusion detection techniques for vehicular communication networks.

2/18/2026 12:00:00 AM

89737

1404/12/01

A Review on Machine Learning-Based Cyber Intrusion Detection Techniques for Vehicular Communication Networks

Automotive Cybersecurity , Intrusion Detection Systems , Machine Learning , Deep Learning; Controller Area Network (CAN)