Byzantine-Resilient Impulsive Control for Bipartite Consensus of Heterogeneous Multiagent Systems

This paper investigates the bipartite consensus problem for heterogeneous multi-agent systems subjected to Byzantine attacks. Byzantine agents send erroneous signals to their neighbors while utilizing incorrect input signals themselves, posing significant challenges for defense. To defend against Byzantine attacks, we propose a resilient heterogeneous impulsive bipartite consensus algorithm for multi-agent systems. This approach ensures that information transmission occurs exclusively at sampling points, significantly reducing control costs, minimizing communication redundancy, and enhancing system robustness. During each sampling event, agents eliminate the most extreme values from their neighbors and utilize the remaining information to generate the control input. By employing this resilient scheme and leveraging the properties of Sarymsakov matrices, we demonstrate that the proposed impulsive control method effectively limits the impact of Byzantine attacks. We also determine the maximum allowable number of Byzantine agents and the corresponding network robustness required to ensure the agents achieve bipartite consensus. Finally, simulations and experiments validate the effectiveness of the proposed approach.