The Optimal Distributed Weighted Least-Squares Estimation in Finite Steps for Networked Systems

This brief investigates the distributed state estimation problem for networked systems. Based on the coupled measurements and information exchanged with neighboring nodes, we propose a distributed estimation algorithm for each node to obtain the state estimate, which is equivalent to the centralized weighted least-squares (WLS) one. When the communication network of the system consists of an acyclic graph, the distributed state estimation algorithm converges in finite iterations. Then we demonstrate that the size of the abstract graph determines the number of iterations. Furthermore, we propose an algorithm for detecting false data attacks by distributed calculation. Finally, the numerical results verify the effectiveness of developed algorithms.