Secure MPC-Based Path Following for UAS in Adverse Network Environment

Abstract

This work considers the path-following problem for Unmanned Aerial System (UAS) in which an online remote control station computes and sends control input signals to the vehicle over an adverse communication network. In that network configuration, the cyber attackers and malicious eavesdroppers are prone to erode the UAS's safety properties such as operational security and information privacy. To guarantee these properties, we introduce a secure model predictive control (MPC) framework for achieving both optimal and safe path-following performance. The unique feature of this framework is that it can simultaneously address all the adversaries occurring in both remote station and network transmission links. Then, an encrypted MPC law is designed using an effective encoding scheme and the Paillier cryptography scheme. It is shown that the closed-loop stability can be guaranteed under the proposed MPC law. Simulation studies of UAS path-following are conducted to validate the effectiveness of the proposed framework.