Learning-based Detection of GPS Spoofing Attack for Quadrotors

TL;DR

The paper addresses GPS spoofing and false data injection attacks on quadrotor UAVs operating under nonlinear, non-Gaussian dynamics, formalized with $x_{k+1} = f(x_k,u_k) + w_k$, $y_k = g(x_k) + v_k$, and attack $d_k$ such that $y_k = g(x_k) + v_k + d_k$. It introduces QUADFormer, a transformer-based detector that combines an EKF-derived residue generator with a semi-supervised learning-based detector and a resilient state estimation module to defend against attacks. Across simulations and real-world tests, QUADFormer achieves higher $F1$-scores than traditional residue-based methods (e.g., CUSUM, SPRT, BHT) and learning-based baselines (e.g., SVM, CNN, LSTM) under varying noise models and attack intensities, demonstrating robustness to non-Gaussian noise and nonlinear dynamics. The framework enhances safe outdoor UAV operation and offers a pathway to extend to broader cyber threats and more advanced residue generation strategies.

Abstract

Safety-critical cyber-physical systems (CPS), such as quadrotor UAVs, are particularly prone to cyber attacks, which can result in significant consequences if not detected promptly and accurately. During outdoor operations, the nonlinear dynamics of UAV systems, combined with non-Gaussian noise, pose challenges to the effectiveness of conventional statistical and machine learning methods. To overcome these limitations, we present QUADFormer, an advanced attack detection framework for quadrotor UAVs leveraging a transformer-based architecture. This framework features a residue generator that produces sequences sensitive to anomalies, which are then analyzed by the transformer to capture statistical patterns for detection and classification. Furthermore, an alert mechanism ensures UAVs can operate safely even when under attack. Extensive simulations and experimental evaluations highlight that QUADFormer outperforms existing state-of-the-art techniques in detection accuracy.

Figures (3)

• Figure 1: UAV operation under GPS spoofing attacks.
• Figure 2: Overview of QUADFormer framework.
• Figure 3: Real-world experiment setup.