Sampled-data control for cyber physical systems via identifying false data injection attacks

In this paper, we discuss the security control problem of cyber physical systems based on identifying false data injection attacks. First, the unknown adversary randomly attacks the measurement output, a mechanism is set in the sensor to update the measurement output to the constant, called the one-valued quantized mechanism. Then, the Bienayme-Chebyshev inequality and the one-valued quantized mechanism are utilized to design an identification algorithm, the estimated value of the false data injection attack and the minimum identification time can be obtained. Furthermore, based on the LMI theory, sampled-data control gains with and without attacks are obtained to ensure the system mean-square exponential stable, respectively. And the corresponding maximum sampling period of the controller is obtained. Finally, numerical examples verify the correctness and effectiveness of the theory.