Privacy-Preserving Fusion Estimation for Power System Based on Channel Gain Encryption*

This paper studies distributed fusion estimation for power system, where smart sensors generate local estimates and send them to fusion center through wireless communication channels in presence of an eavesdropper. To protect power system state privacy, the privacy-preserving method based on artificial noise technique is introduced, which is designed by using channel gain encryption for each sensor. A sufficient condition on artificial noise is provided to make the eavesdropper's expected error covariance unbounded for power system while the user's fusion estimator will not deteriorate. Numerical example of single machine infinite bus power system is presented to verify our results.