Resilient distributed state estimation with multi-hop communication

Abstract

This paper investigates the problem of resilient distributed state estimation for a linear system in multi-hop communication networks. First, we introduce a graph-theoretic property, called multiple-hop strong robustness, to formalize the notion of communication and measurement redundancy in multi-hop communication networks. Second, we propose multi-hop communication-based resilient distributed state estimation strategies in the synchronous network and the asynchronous network with time delays, respectively. We provide necessary and sufficient conditions on the network topologies under which the regular agents can still asymptotically estimate the state vector of the system despite the existence of adversarial agents using the property of multiple-hop strong robustness. In contrast to the 1-hop communication-based strategies in the existing results, the proposed multi-hop communication-based strategies can reduce the communication redundancy requirement on the network and accelerate convergence. Finally, we present two examples to verify the effectiveness of the proposed strategies.