Attack-Resilient Distributed Control of Multi-Agent Systems Under Output Feedback-Driven Multi-Channel DoS Attacks

This paper investigates the distributed resilient output-tracking control problem for heterogeneous multi-agent systems (MASs) subject to multi-channel denial-of-service (DoS) attacks. First, we propose a simplified framework to address the distributed output tracking control for systems with uncertain dynamics. Then, we develop a distributed output feedback-driven resilient control protocol under asynchronous multi-channel DoS attacks, considering both inter-follower communication channels and leader-follower links. To enhance resilience against attacks, a set of channel-dependent decay rates is introduced and designed to ensure self-healing output tracking capabilities under active attacks. Unlike conventional distributed output control methods that heavily depend on distributed observers and assume the leader’s state is globally accessible or transmitted via distributed communication, our resilient control algorithm eliminates the need for distributed observers while reducing communication overhead through compact output information encoding. Finally, the effectiveness of the proposed algorithm is demonstrated through a practical example involving a multi-vehicle system. Note to Practitioners—For practical systems such as ground vehicles, aerial vehicles, and robots, the dynamics often exhibit heterogeneity and uncertainty due to real-world operational conditions. To ensure robust tracking performance under such challenges, this paper investigates resilient secure tracking control for heterogeneous and uncertain multi-agent systems (MASs) under distributed denial-of-service (DoS) attacks. Unlike existing methods that rely on distributed observers requiring the leader’s full state, we propose a novel output feedback-driven resilient control framework for heterogeneous and uncertain systems to address challenges arising from model uncertainty and limited information availability in practice. The proposed distributed resilient control algorithm is applicable to formation control of autonomous ground vehicles, aerial vehicles, and intelligent robots. Its advantages lie in avoiding the transmission of observed variables and enabling lightweight communication. Additionally, an attack immunity mechanism is integrated to counter multi-channel DoS attacks. A set of equivalent decay rates is defined to provide sufficient conditions for secure tracking. Finally, the effectiveness of the proposed algorithm is demonstrated through a practical example involving a multi-vehicle system.