Abstract
The growing use of Industrial Internet of Things (IIoT) technologies in transportation infrastructure increases the system's exposure, making them more vulnerable to cyberattacks. As a key component in modern vehicles, Controller Area Network (CAN) systems enable the real-time detection of ongoing intrusions. However, CAN networks are still at risk from various types of attacks, such as Denial of Service (DoS), Fuzzy attacks, Gear Spoofing, and RPM Spoofing. This paper presents an attack detection and simulation model for IIoT-enabled vehicular systems that exploits Edge Computing. In this work, the CAN bus network-related cyberattacks such as DoS, Fuzzy, Gear Spoofing, and RPM Spoofing are identified by using XGBoost and LR models. Experiments conducted on the Car-Hacking Dataset indicate good classification accuracy reaching 96%. Cognitive system for design and is inspired by the C-OODA (Сognitive-Observe, Orient, Decide, Act) concept for cognitive engineering.