Efficient Algorithm for Secure State Estimation Using Adam Optimizer and ADMM

This paper addresses the challenge of efficient secure state estimation of large-scale cyber-physical systems modeled by linear Gaussian systems with compromised sensors. We deal with this issue by providing an efficient implementation method based on the Adam optimizer and the alternating direction method of multipliers for a secure state estimation method based on fusion of multiple local estimators. We then explore scenarios involving a two-state system and cases where the number of system states or sensors increases. Simulation results demonstrate that the proposed algorithm maintains robust performance and scalability across different configurations, achieving a balance between computational efficiency and estimation accuracy.