Compensation of Linear Attacks to Cyber Physical Systems through ARX System Identification

Nowadays, Cyber-physical Systems (CPSs) are playing important roles in critical infrastructures. In these systems, using computer networks for communication is quite common, but this makes them vulnerable to cyber-attacks. In this paper, we design an intrusion detection and compensation algorithm through a system identification framework. At first, an intrusion detection approach is proposed which detects anomalies in the output of the physical process (plant) by comparing it with the output of a reference model which is built locally. Once an attack is detected, the process of identifying its effective function begins. Then, a new controller is designed on the fly to compensate the effects of attack and keep the system stable. In the paper at hand, a DC motor has been used as the plant in simulations. The control system and the proposed detection and compensation framework have been simulated by using Matlab Simulink. Results show that the proposed framework can well detect and control the malicious effects of some deception attacks on the forward channel of the simulated system