Secure Cooperative Path Following of Autonomous Surface Vehicles Under Cyber and Physical Attacks

Abstract

This paper is concerned with the networked cooperative path following (CPF) problem for multiple autonomous surface vehicles (ASVs) subject to simultaneous cyber and physical attacks. First, to compensate the adverse effects of the physical-attack-induced bias injections, an extended state observer is designed to provide real-time estimates of the unmeasured velocities and unknown nonlinear terms. Next, to identify and handle various cyber attacks, a novel secure data transmission mechanism, featuring a secure transmitter and a secure receiver, is developed for each ASV. Then, a secure CPF control scheme, consisting of a networked cooperative kinematic control law and a networked kinetic control law, is presented. Furthermore, the observer error dynamics and networked CPF error dynamics are derived to account for the simultaneous network-induced delays, packet dropouts, physical attacks, and cyber attacks. The proposed control scheme is capable to preserve satisfactory secure tracking performance of the resulting CPF control system under a desired reference path even in the presence of external environmental disturbances, delays, packet dropouts, and malicious attacks. Finally, several case studies are provided to substantiate the effectiveness of the secure CPF control scheme.