Improving Kalman filter for cyber physical systems subject to replay attacks: An attack-detection-based compensation strategy

This paper investigates the problem of improving Kalman filtering for a class of multi-layer cyber physical systems (CPS) in the face of multiple consecutive replay attacks. CPS is vulnerable in an open network environment, and the replay attack model is described based on the number and duration of successful message transmissions. For all possible occurrences of replay attacks, a detection and compensation method is obtained with the help of the residual monitoring mechanism and an attack detection function. At the same time, in order to identify the attack has occurred in which, the register is integrated into the system. Furthermore, a novel Kalman filter with a compensation mechanism is designed to estimate the ideal states of CPS subject to replay attacks, and the Kalman filter gain is optimized based on the detection results. Finally, a simulation example is given to verify the effectiveness of the proposed improving Kalman filter.