Sequential cyber-attack detection in the large-scale smart grid system

This paper investigates the sequential detection of cyber-attack in smart grid system, which undermines the power system state estimation by injecting malicious data to the monitoring meters. To overcome the challenges raised by unpredictable attack features such as injected data information and the set of affected meters, we propose a generalized Cumulative Sum (CUSUM) detector based on the generalized likelihood ratio. Furthermore, to alleviate the exponentially growing computational burden, we further refine the detector such that its computational complexity scales linearly with the number of meters, which is usually large in the smart grid system.