Malicious data injection attack against power system state estimation based on orthogonal matching pursuit

State estimation is a critical power system component that estimates the state of the power network and deals with bad data, depending in general on a redundant set of meter measurements and network topology configuration. Recently, some researchers have constructed a new class of attack which can successfully bypass the existing power system state estimation and inject bad data to the state variables, causing enormous threats to the power system. This paper investigates the methods to identifying the minimum number of meter measurements to compromise in launching such an attack, which is named least-effort malicious data injection attack. A modified orthogonal matching pursuit (OMP) algorithm is introduced here for identifying the meters, since traditional matching pursuit (MP) algorithm requires a large number of iterations to reach convergence. Comparison of the two methods in the simulation on standard IEEE test system indicates that the OMP algorithm compromises fewer meters than the MP algorithm in the same number of iterations.