Controller Area Network

Controller Area Networks (CANs), the primary communication protocol in modern vehicles, are increasingly vulnerable to cyberattacks due to their inherent lack of security features. Current research focuses on developing robust intrusion detection systems (IDSs) using diverse machine learning approaches, including deep learning models (like transformers and autoencoders), graph machine learning, and one-class classification methods, to identify both known and unknown attacks, particularly sophisticated masquerade attacks. These efforts are crucial for enhancing automotive cybersecurity and ensuring the safety and reliability of autonomous driving systems, with a strong emphasis on improving detection accuracy and reducing false positives.