Towards a Self-rescuing System for UAVs Under GNSS Attack

TL;DR

This work tackles UAV vulnerability to GNSS disruption by proposing a lightweight, GNSS-denied navigation framework that explicitly accounts for wind. It introduces a two-phase approach: Forward Phase for inertially tracking the outward path and Backward Phase for generating a wind-aware return route, using either a Weighted Proportion Method or an ML-based LASSO regression to adapt speeds under wind influence. Evaluation on indoor and outdoor flights with a Mavic 3 Classic and an anemometer demonstrates the ability to return to the takeoff point under wind, achieving mean position errors on the order of tens of meters and inference times in the low-millisecond range. The results highlight practical trade-offs: the Weighted Method is simple and fast but sensitive to wind changes, while the ML-based method offers improved robustness at the cost of offline training and model maintenance, with potential for real-time onboard deployment in future work.

Abstract

There has been substantial growth in the UAV market along with an expansion in their applications. However, the successful execution of a UAV mission is very often dependent on the use of a GNSS. Unfortunately, the vulnerability of GNSS signals, due to their lack of encryption and authentication, poses a significant cybersecurity issue. This vulnerability makes various attacks, particularly the "GNSS spoofing attack," and "GNSS jamming attack" easily executable. Generally speaking, during this attack, the drone is manipulated into altering its path, usually resulting in an immediate forced landing or crash. As far as we know, we are the first to propose a lightweight-solution that enable a drone to autonomously rescue itself, assuming it is under GNSS attack and the GNSS is no longer available, and return safely to its initial takeoff position, thereby preventing any potential crashes. During the flight, wind plays a critical role as it can instantaneously alter the drone's position. To solve this problem, we have devised a highly effective 2-phases solution: (i) Forward Phase, for monitoring and recording the forward journey, and (ii) Backward Phase, that generates a backward route, based on the Forward Phase and wind presence. The final solution ensures strong performance in consistently returning the drone to the original position, even in wind situations, while maintaining a very fast computation time.

Figures (8)

• Figure 1: Whole set up utilized: (I) is the USB Adapter, (II) is the Trisonica Anemometer, (III) shows the Raspberry PI 3 and, (IV) is the power for Raspberry.
• Figure 2: It is shown the same route of a flight in a 2D and a 3D space.
• Figure 3: The same flight but displayed with two different methods.
• Figure 4: LASSO regression plot for both cases.
• Figure 5: Backward Phase Evaluation with Weighted Proportion Method ( Section \ref{['sec:backward']} and algorithm \ref{['alg:backpropagation']}). Same flight plotted using different value of $\alpha$ and $\beta$.