Resilient Distributed Quadrotor AAVs Game: Addressing FDI and Physical Attacks

Abstract

Distributed quadrotor autonomous aerial vehicles(AAVs) are increasingly deployed in critical applications, including reconnaissance, disaster response, and transportation, where system resilience is paramount. However, their complex dynamics and reliance on cyber-physical systems make them highly vulnerable to adversities such as uncertain dynamics, time-varying input disturbances, false data injection (FDI), and physical attacks. This paper proposes a hierarchical framework consisting of a game layer and an execution layer to enhance the resilience of a noncooperative game involving distributed AAVs. In the game layer, a novel FDI observer mitigates communication and computational FDI attacks, facilitating convergence to infinitesimal ranges around the Nash equilibrium (NE) point. The execution layer addresses uncertain dynamics and physical attacks, ensuring that AAVs can track a narrow neighborhood of the NE even in the presence of actuator failures. Extensive simulations validate the algorithm's effectiveness and robustness against mixed attack scenarios, underscoring its potential for real-world AAV applications.