Cyber security impacts on all-PMU state estimator - a case study on co-simulation platform GECO

Traditional state estimators require longer scan time, leading to delayed, and inaccurate state estimation. Given the increased deployment trend of phasor measurement units (PMUs), it is expected that all-PMU state estimation will eventually replace traditional or mixed state estimators at the control centers of power utilities. Due to the repeated calibration of the voltage and current transformers at the measurement sites, and direct time-synchronized measurement of phasors, the estimated state by an all-PMU state estimator is not only accurate, but also available at a rapid rate, leading to the use of the system state for protection, stabilization, and even calibration of the measuring devices. However, due to high reliance on an advanced communication network infrastructure for the delivery of large amount of measurements in real-time, the cyber attack surface of the power system is increased. Deliberate cyber attacks or unintentional network failures can affect the state estimator leading to misoperations of the power system. In this paper, we study the cyber security impacts on the all-PMU state estimator, using a power system and data network co-simulation method. A linear state estimator for a model of the New England 39-bus system and the corresponding data network is built in a global event-driven co-simulation platform “GECO” which was developed and leveraged for our experimental setup. The co-simulation of PSLF (power system simulator) and NS-2 (network simulator) is run with injection of attacks on the network. The injected cyber attacks in the form of network failures or malicious data injection are simulated and their effects are observed. We also, observe the robustness of the all-PMU state estimator, when the number of affected measurements is below a threshold.