Attack prevention and detection for cyber-physical systems based on coprime factorization technique

While cyber-physical systems (CPSs) bring convenience to industrial production and human life, they are, in the mean time, exposed to the threat of cyber-physical attacks, which have the characteristics of concealment and vital destruction. Recent reports on several horrible attack events aroused people's attention to the security of CPSs. Motivated by this, this paper proposes a defense scheme based on co-prime factorization technology, consisting of attack prevention and detection. This scheme utilizes residual signals that are commonly used in fault diagnosis to realize secure transmission because these residual signals do not contain essential system knowledge. At the same time, a filter module is added, which makes use of noise signals to cover and protect the reference signals and the operation status of the system. By this means, the secure transmission method reduces the leakage of system information and interferes with the design of stealthy attacks, which need sufficient system knowledge. Furthermore, it realizes the detection of non-stealthy attacks under the same framework. Simulation results on a linear system with random parameters demonstrate the effectiveness of the proposed scheme.