Distributed Secure State Estimation Against Sparse Attacks: A Sensor Clustering Approach

This paper investigates the problem of sparse attack isolation and distributed secure state estimation (SSE) for remote sensor network in an attack-prone environment. The attack considered in the present paper is a variant of the sparse sensor attack model, and is characterized by the number of sensors clusters that can be compromised at each moment. By analyzing the worst-case scenario in the presence of the considered attack, an attack detection and isolation mechanism based on the weighted communication design is provided to block the spread of the attack across the observer network. Moreover, based on the attack isolation, a distributed SSE algorithm is constructed to ensure globally consistent state estimation. Finally, a numerical simulation is provided to validate the effectiveness of the proposed conditions and algorithm.