Dynamic Attack Path Prediction and Visualization for Industrial Cyber-Physical Systems Under Cyber Attacks

The accurate and effective prediction of network attack paths has become a crucial concern in the realm of network security, given the inherent uncertainty and subjectivity associated with network attack methods. To solve this problem, this paper proposes a visualized dynamic attack path prediction scheme for industrial cyber-physical systems (ICPSs). The method combines the Bayesian attack graph with the knowledge graph and considers the topology of the digital twin layer to make it closer to the actual situation. In addition, node dynamic reachability probabilities are considered to provide support for the interpretation of the prediction results. The simulation results demonstrate that the proposed scheme is more flexible and scalable than the static attack graph. These improvements enable more accurate prediction of the network attack path and enhance the network's security protection ability.