Robust Covert Attack on Cyber-Physical Systems Based on Adaptive Control

Widespread application of network in control systems has significantly increased the risk of cyber attacks. In this paper, we have investigated a specific class of cyber attack in which the adversary has partial model information of the plant. The adversary can make his attack stealthy by applying one of the well-known adaptive control approaches called model reference adaptive control (MRAC). This kind of robust covert attacks can be very dangerous as the adversary can take over the control of the plant through the communication network without being noticed by the monitoring system. A simulation example is used to illustrate the risk of this kind of robust stealthy covert attacks.