A method for modeling and evaluation of the security of cyber-physical systems

Abstract

Quantitative evaluation of security has always been one of the challenges in the field of computer security. The integration of computing and communication technologies with physical components, has introduced a variety of new security risks, which threaten cyber-physical components. It is possible that an attacker damage a physical component with cyber attack. In this paper, we propose a new approach for modeling and quantitative evaluation of the security of cyber-physical systems (CPS). The proposed method, considers those cyber attacks that can lead to physical damages. The factors impacting attacker's decision-making in the process of cyber attack to cyber-physical system are also taken into account. Furthermore, for describing the attacker and the system behaviors over time, the uniform probability distributions are used in a state-based semi-Markov chain (SMC) model. The security analysis is carried out for mean time to security failure (MTTSF), steady-state security, and steady-state physical availability.