Attack-resilient measurement design methodology for State Estimation to increase robustness against cyber attacks

Traditional measurement design problem in State Estimation provides a strategy for optimal placement of meters to ensure the overall system observability under possible single contingencies and loss of measurements. However, with the increased interconnectivity and exposure to public network infrastructures, the SCADA network is highly vulnerable to sophisticated cyber attacks that could potentially cause loss of meter data integrity or availability across multiple substations. It has been shown that the state estimates can be biased by malicious adversaries through carefully chosen topology manipulations or false data injections. State Estimators typically use various approximate pseudo-measurements to handle cases where SCADA data does not provide total system observability. However, a loss of State Estimator accuracy when the system is operating under severe stress could cause the operator to take incorrect decisions. Therefore, it is extremely important to consider the system operating conditions, possibility of several types of contingencies and loss of measurements beyond existing assumptions in the measurement design problem to make the state estimator resilient to different types of cyber attacks.