Stealth cyber-attacks impact mitigation on UAVs sensors using multiple model adaptive control technique

Abstract

This thesis laser focused on how to secure UAV systems against stealth cyber-attacks. In-depth literature review has been conducted to understand UAV vulnerabilities and the current techniques that have been proposed or used to tackle cyber-attacks. Then, details of various types of cyber-attacks have been presented based on their strategy methods, assuming unsecured communication channels and successful injection of false data into UAV sensors. The main challenge in securing UAV against cyber-attacks is to diagnose the attacks systematically as they are in many cases are unpredictable and appear in erratic ways.
Stealth cyber-attacks have the ability to modify sensor readings by tampering with individual meters through the communication channels of the UAV. This can cause delays and congestion in the communication channels, which leads to significant interruptions to its behaviour. Strategies involving traditional software security, such as encrypted communication and memory isolation, are insufficient to protect against sensor tampering and artificial delay attacks. Therefore, securing UAV systems against cyber-attacks is still a gap in the literature and the research community.
This research project tackles the demanding UAV cyber security challenges and describe the use of a novel possible solution of adaptive secure control system to mitigate impacts on UAV sensors from stealth cyber-attack, in particular Denial of Service (DoS) and False Data Injection (FDI). The design of adaptive secure control system has been achieved by utilising Multiple Module Adaptive Control (MMAC) technique and developing a bank of Kalman Filter (KF) models and Controller Bank with PID controllers to mitigate the impacts in different scenarios and drive the UAV system to the desired behaviour. Switching algorithm has also been developed and designed to guide the transition between the controllers based on each KF model weighting. MATLAB / Simulink have been used to build and simulate the whole system including DoS and FDI stealth cyber-attacks. The simulation results show that the design of the developed adaptive secure control system in this project enhanced the system performance and successfully mitigated cyber-attacks impacts on Fixed Wing UAV and Quadcopter UAV.

More information

Identifiers

ORCID for Suleiman Sharkh: orcid.org/0000-0001-7335-8503

Catalogue record

Export record