Complementarity reformulations for false data injection attacks on PMU-only state estimation

The accuracy and integrity of state estimators are perniciously influenced by false data injection attacks (FDIAs) trying to manipulate the values of a subset of measurements without being detected by the underlying bad data processing scheme. In the framework of vulnerability analysis, we examine FDIAs against a PMU linear state estimator based on Cartesian formulation in the presence of zero injection buses, under the assumption that the attacker would most likely try to corrupt as few measurements as possible. Exact and relaxed complementarity reformulations to cardinality minimization are proposed in order to compute minimal sets of measurements whose values need to be maliciously modified for successful FDIAs. The performance of the proposed approaches is demonstrated with experimental evaluations over IEEE benchmark systems.