Research

Intrusion Detection Systems (IDS) for Controller Area Networks (CAN) Bus

Our research focuses on improving the security of Controller Area Networks (CAN) in modern vehicles, which are increasingly dependent on interconnected electronic control units (ECUs) for critical functions. The CAN protocol, designed for reliability and efficiency, lacks built-in security measures, making it vulnerable to sophisticated cyberattacks that exploit these gaps. Traditional IDS methods, which monitor timing and message sequences, are inadequate against advanced attacks that manipulate signal semantics. Our work addresses these challenges by developing detection systems that can identify such subtle intrusions, enhancing the overall security of vehicular networks.

Robust Misbehavior Detection System (MBDS) For Vehicle-to-Everything (V2X) Communication

Our research addresses the pressing need for robust misbehavior detection systems (MBDS) in Vehicle-to-Everything (V2X) communication networks, which are essential for modern intelligent transportation systems. V2X networks face significant security challenges due to their open, decentralized nature, which exposes them to advanced adversarial attacks that can evade traditional machine learning-based detection systems. To overcome these issues, we develop Vehicular GAN (VehiGAN), an advanced MBDS leveraging Generative Adversarial Networks (GANs). VehiGAN enhances anomaly detection and offers strong resilience against adversarial manipulations, improving the security and reliability of V2X communications..