Design Of False Data Injection Attacks For Output Tracking Control Systems

A cyber-physical system(CPS) is a system of collaborating computational elements and physical entities. CPS has a diverse group of application fields, including environmental monitor and control, intelligent transportation, smart grid, defense message system, critical infrastructure(e.g., electricity, communication, water and gas), intelligent robot, aerospace information security and many other systems vital to our well-being. With the development of CPS, the security becomes an important and challenging problem. In the CPS, the control commands and sensor measurements are transmitted over communication channels which are susceptible to being corrupted by adversaries. Attackers can launch various attacks, including false data injection attacks, replay attacks and denial of service(DoS) attacks, to destroy the system performance or even destabilize the system without being attacked. In this paper, we study the false data injection attacks in the area of secure control theory, aiming at destroy the control system without being attacked.In this paper, the output tracking control systems are investigated, in which the dynamics is modeled as a discrete-time linear time-invariant Gaussian control system. By including the output tracking error as an additional state, an augmented state is designed based incremental state and additional state, which lead to an augmented control system with the incremental output. In the output tracking controller, Kalman filter-based incremental state observer and linear quadratic Gaussian-based augmented state feedback control strategy are designed. The Euclidean-based detector is used for detecting the false data injection attacks. We designed two false data injection attacks, include false data injection attacks for sensor measurements and false data injection attacks for sensor measurements and control commands, which based on the communication channels over the controller and the sensor, the controller and the actuator. All we designed false data injection attacks can destroy the control system without being attacked. Two stability analyses and simulation results are employed to illustrate the effectiveness of the designed false data injection attacks. To further verify the designed false data injection attacks for control commands and sensor measurements on practical systems, a brushless DC motor-based Internet control system test rig has been built.