Distribute Nash Equilibrium Seeking for Networked Agent Games with Time-Varying Communication Constraints

In this paper, we explore the challenge of seeking the Nash equilibrium(NE) in time-varying agent networks. Considering the impact of dynamic communication network with Denial-of-Service (DoS) attacks, we propose a distributed control strategy that ensures agents convergence to the NE despite intermittent communication disruptions caused by DoS attacks. The method utilizes a virtual observer to represent the dynamics of the real agents and applies a gradient-based update rule to guide the system towards equilibrium. The theoretical analysis shows that under weight-balanced and strongly connected communication topologies, the system can still exponentially converge to the equilibrium point as long as the duration of the DoS attack satisfies a certain proportion condition. Numerical simulation involving a multi-agent system (MAS) confirm the proposed method's effectiveness, demonstrating that the agents' states reach the NE even in the presence of communication disruptions.