Distributed state estimation under denial of service

This paper investigates the properties of consensus-based distributed state estimation (DSE) in the presence of denial of service (DoS) attacks. The considered model characterizes the DoS attack in terms of its average frequency and duration. It is shown that a family of DSE algorithms, namely Hybrid Consensus on Measurements - Consensus on Information, preserves its stability properties under DoS provided that a sufficiently high transmission rate is ensured depending on the bounds on DoS frequency and duration. The theoretical findings are confirmed by numerical simulations in a target tracking case study.