Secure remote state estimation against linear man-in-the-middle attacks using watermarking

In this paper, an attack defense method is proposed to address the secure remote state estimation problem caused by linear man-in-the-middle attacks in cyber-physical systems (CPS). We utilize the pseudo-random number as a watermarking to encrypt and decrypt the data transmitted through the wireless network. Via the proposed method, the transmitted data in the normal operation can be recovered. Since the data modified by the attacker can be marked with the watermarking, the chi(2) detector is capable of detecting the attack. For three different attack scenarios, we analyze the evolution of the remote estimation error covariances and the detection performance, respectively. In the sense of minimizing the estimation error covariance, the optimal parameter set of the watermarking is derived. Furthermore, the proposed method can even be extended to detect the false data injection attack and the replay attack. Finally, several examples are provided to illustrate the derived results. (C) 2020 Elsevier Ltd. All rights reserved.