A Periodic Dynamic Encoding Mechanism for Stealthy Attack Detection in Distributed State Estimation

ABSTRACT This article deals with the design and defense issues of false data injection attack in distributed state estimation. At each time step, each sensor sends its estimate to neighboring sensors. To verify the security of the received data, a column vector composed of state estimate differences is used as the detection variable. Based on the data directly obtained by the attacker in the wireless communication network, an attack strategy is established to degrade the system estimation performance while not being detected by the detector. Under the attack strategy, the update of the estimation error covariance is analyzed and the optimal attack parameters are obtained. To enhance the security of the system, a periodic dynamic encoding mechanism is proposed to prevent attackers from eavesdropping on original data, which consists of two stages: preparation stage and periodic encoding stage. It is proved that the proposed encoding mechanism can destroy the stealthiness of the attack, thus ensuring the detection ability of the detector. Numerical simulations are given to verify the results of the theoretical analysis.