Deception Attack Detection and Estimation for a Local Vehicle in Vehicle Platooning Based on a Modified UFIR Estimator

Abstract

In this article, the position sensor deception attack detection and estimation problem is investigated for a local vehicle in a vehicle platoon. In a platoon system, the position measurement is critical as the distances between neighboring vehicles are relatively small. However, the position measurement of vehicles is usually vulnerable to deception attacks as it relies on external information, such as GPS and environment information from cameras. Therefore, position sensor deception attack detection and estimation should be addressed for local vehicles in a platoon. To deal with this problem, a linearized model is presented to describe the longitudinal dynamics of a local vehicle. Moreover, modeling uncertainties, measurement noises, and piecewise constant deception attacks injected in position measurement are specified along with this model. Based on this model, a scheme based on a modified unbiased finite impulse response (UFIR) estimator is proposed to generate an intermediate estimated value related only to the attack. Then, the deception attack is recovered based on this value through a function fitting strategy. Based on analysis results, simulations are conducted to verify the effectiveness of the proposed attack detection and estimation scheme.