Observer Design for Autonomous Systems Under Sensor Attacks

As autonomous systems become more widespread, they are increasingly vulnerable to sensor attacks. In this paper, we propose a novel observer designed to estimate the magnitude of sensor attacks on autonomous systems. Our proposed observer is a discrete-time observer that can be used for joint estimation of both state and parameters. The key idea behind our proposed observer is to filter the measurement and augment the measurement equation into the state space model. Our proposed observer has several advantages over existing methods. First, the filtering of the measurement and augmentation of the measurement equation enable the observer to account for the impact of sensor attacks on both the state and the parameters of the system. Second, the joint estimation of state and parameters enables the observer to adapt to changing conditions and maintain accurate estimates over time. We demonstrate the effectiveness of our proposed observer through numerical simulations. Our results show that the observer is capable of accurately estimating the magnitude of sensor attacks on autonomous systems.