Attack detection and security control for quadrotor under stealthy attacks

Abstract

When malicious attacks occur, the security issues of unmanned aerial vehicles are full of challenges. This paper investigates the attack detection and security control problems for quadrotors with stealthy false data injection attacks. Decoupled linear models are provided to describe the dynamics. A novel cyber-attack detection scheme integrating watermarking signals and auxiliary functions is devised to detect stealthy false data injection attacks. To determine which sensor data is attacked, an attack location algorithm is developed based on the observability of the derived linear model. Furthermore, a security control scheme is provided to mitigate stealthy false data injection attacks. Finally, simulation results are presented to validate the proposed theories in this paper.