Dynamic Information Security Risk Assessment for Railway Signaling Safety Data Network Based on Bayesian Inference

The railway signaling safety data network is a dedicated network for transmitting railway signaling safety information in CTCS-2/3 train control systems. Security attack may cause equipment failure or even safety accidents. Therefore, dynamic perception of its security state is required during operation. This paper proposed a dynamic information security risk assessment method based Bayesian inference for railway signaling safety data network. First, the logical relationship between multi-step attacks and the influence propagation path were analyzed, followed by the establishment of an extended Bayesian attack graph model based on the uncertainty of the atomic attack. Next, based on the correlation between attacks and detection equipment alarms, real-time observed attack information was used as evidence for the dynamic reasoning of accident probability. Finally, the system risk was obtained by combining the functional safety impact of the accident. Simulation experiments verify the effectiveness of the method for dynamic risk assessment. It can be used for signaling safety data network risk analysis and provide a basis for the security mitigation scheme. © 2020, Department of Journal of the China Railway Society. All right reserved.