Resilient Control for Multi-Agent System Based on Barrier Lyapunov Functions Under Replay Attack

This paper addresses the resilient leader-follower tracking control problem for nonlinear MASs subject to replay attacks. Unlike existing methods that primarily focus on stability, our objective is to simultaneously ensure stability and state constraints satisfaction even during attacks. A distinct contribution of this work lies in the quantification of safety resilience: namely, we derive a condition explicitly characterizing the tolerable limit of replay attacks relative to the system's safety bounds. This resilience condition guarantees that the safety barrier remains intact. Simulation results validate that, under the derived condition, the proposed distributed controller guarantees that all follower outputs asymptotically converge to the leader's trajectory while strictly remaining within the preassigned safe bounds.