Abstract—In the attack of power Cyber Physical System (CPS), the key problem of taking effective response to the defense measures is how to extract the attack path quickly and accurately. This paper proposes an attack path extraction model based on Hidden Markov Model (HMM). Firstly, the original state sequence of information communication system and electric system must be processed by the representing, filtering, segmenting and fusion, then receiving the joint sequence of system failure that produced by the same attacker; Secondly, on the basis of the designed mapping table between information physics cooperative attack and system failure, system failure probability matrix can be generated dynamically. Meanwhile, the concept of sensitivity matrix is introduced to quantitative analyze the interaction effects between information attacks and physical attacks. Finally, the implementation algorithm of the model is given. The experimental results show that the proposed model can effectively extract the most likely sequence of attack path based on the known sequence of the system failure.

Index Terms—Power cyber physical system, attack path extraction, the joint sequence of system failure, hidden markov model

Cite: Lei Wang, Zhaoyang Qu, and Zelong Li, “The Design and Implementation of Attack Path Extraction Model in Power Cyber Physical System," Journal of Communications, vol. 11, no. 9, pp. 834-840, 2016. Doi: 10.12720/jcm.11.9.834-840