Intelligent connected vehicles (ICVs) integrate advanced communication, sensing, and computing technologies to significantly enhance traffic efficiency, safety, and driving experience. However, its highly interconnected cyber-physical system (CPS) characteristics render it vulnerable to cyber threats, such as false data injection (FDI), denial of service (DoS), and spoofing attacks. This paper systematically reviews the research status and key technologies of ICV cyber-physical security. Firstly, vulnerability analysis, risk assessment and test verification techniques for ICV are reviewed, and attack paths and their potential impacts on controller area network (CAN) bus, vehicle-to-everything (V2X) communication and power systems are analysed. Secondly, it summarizes attack detection methods based on vehicle dynamics, machine learning, and network behaviors, covering autonomous driving, in-vehicle networks, and charging system scenarios, demonstrating high-accuracy real-time detection capabilities. Furthermore, it explores control strategies based on attack separation, disturbance compensation and machine learning, which mitigate the impact of attacks to maintain path tracking accuracy, platoon stability, and powertrain performance. Through comprehensive analysis and technical synthesis, this paper provides a reference for the study and application of ICV cyber-physical security, and envisions future directions in enhancing ICV security through cross-domain scenarios, multi-source information fusion, and artificial intelligence algorithms.
He et al. (Mon,) studied this question.