LAUSR

In this article, we study the problem of resilient consensus for a multiagent network, where some adversarial nodes attempt to prevent consensus of nonfaulty nodes by transmitting faulty values. Our approach is based on that of the so-called mean subsequence reduced (MSR) algorithm with a special emphasis on its use in agents capable to communicate with multihop neighbors. The MSR algorithm provides an effective technique for agents to achieve resilient consensus if the multiagent network satisfies certain connectivity requirements. Our analysis highlights that for maintaining the same level of resilience against adversarial nodes, such network connectivity requirements can be relaxed by increasing the number of relay hops. In particular, we characterize tight network structures for our algorithm to succeed and propose a novel notion of graph robustness with multihop communication. Moreover, we analyze the multihop W-MSR algorithm with delays in communication since messages from multihop neighbors may require different numbers of time steps for their transmissions. Numerical examples are also presented to verify the efficacy of the proposed method.