Event-based Attack Tolerant for A Class of False Data Injection in Networked Control Systems

This paper addresses an event-based false data injection (FDI) tolerance problem for networked control systems (NCSs). The NCSs are said to be tolerable for attacks, if some extent of state divergence caused by attacks can be tolerated for the studied system. Under the scenario that the measurement (sensor-to-controller) channels are jammed, the average duration of NCSs is well characterized under consideration of the FDI attacks; and the characteristic of m-tolerant for NCSs is refined, according to the specific FDI under an event-triggered control scheme. Moreover, the effects of these FDI attacks to the evolution of systems are deeply analyzed. Examples are given to show the secure feature of the proposed method.