Design And Detection Of False Data Injection Attacks Against Networked Control Systems

Cyber-physical system(Cyber physics system,CPS)has received widespread attention as a new generation of intelligent systems.As more and more areas applying the technology related to the cyber-physical system,the security problems of CPS become particularly important,and how to guarantee the security of cyber-physical system has been the focus of our researchers study,it is also the prerequisite for future development of cyber-physical system.Considering the concept of cyber-physical system is based on networked control system which is characterized by distributed and network communication.For this feature,the attacker can easily get the data from transmission network,and system security is extremely vulnerable.In this paper,we consider the security of control system and study the effect of false data injection attacks on Cyber-physical system,consider an output tracking control system dynamic model is linear time invariant discrete Gauss control system,using the tracking error and the system state to construct an augmented control system.First of all,from the attacker's point,imagine that an attacker could steal transmitting data of the target system in the network,and through the analyzing of the data model of the target system by using subspace identification method,can directly get the exact model of the system.The attacker uses State-space models to design a secret false data injection attack sequence,aims to attack the target and the reference signal of the controlled plant will deviate from the given and detector based on residuals on the controller side cannot detect attacks by false senser data.According to this kind of attack,devised a method to be able to detect these false attacks.Finally through simulations and experiments to verify the effectiveness of attack and detection method.