Improved RFI Localization Through Aircraft Position Estimation During Losses in ADS-B Reception

Michael Dacus, Zixi Liu, Sherman Lo and Todd Walter

Abstract:	The ability to estimate aircraft position and velocity is vital for safe and effective air traffic operations and management. With the emergence of the Automatic Dependent Surveillance-Broadcast (ADS-B) network, aircraft are now capable of sending GNSS positioning and operational data to ground stations to be used by regional air traffic controllers. In the research community, ADSB has also been used for quickly detecting and even more recently, localizing and characterizing GNSS interference threats [1]. However, ADS-B does present some limitations that inhibit GNSS Interference, Detection and Localization (IDL) capability, specifically the loss of regular ADS-B derived positions. Reducing this loss of real-time aircraft positioning information using various interpolation approaches could help immensely with localization. This paper uses historical crowdsourced ADS-B data to examine how well missing flight path data can be interpolated, and how interpolated flight paths can be used to localize interference sources. Crowdsourced data is retrieved from OpenSky Network, a client-based application that receives air traffic data from a network of ADS-B receivers around the globe [2]. ADS-B flight paths are processed, reconstructed, and split into various datasets based on the change in the Navigational Integrity Category (NIC) and presence of gaps in the flight paths. The two primary interpolation methods tested are a simple cubic spline and a MATLAB Systems Object, geoTrajectory. Both were applied to randomized trajectories and aircraft on a nominal test set to understand how well each performed compared to its original ADS-B position. Once interpolation criteria are established for each flight profile, the chosen interpolation method is then applied to a dataset with flight gaps indicating a drop in the NIC value (potential GNSS/RFI interference), as well as another dataset [with flight gaps], but no dramatic change in NIC (possibly beyond LOS). An estimated centroid location of interference is calculated using only the flight paths indicating possible GNSS/RFI interference. To assess the performance of each interpolation method and how it might improve IDL efforts, data was selected over a period of 3 days at a reported interference event near Denver International Airport [3]. Based on the results from the various interpolation methods, the nominal data showed that both cubic spline and geoTrajectory had similar performance, and that more complex flight path geometries are harder to interpolate. Once the interpolation method is applied to the gapped flight datasets, predicted flight paths from data with indicated drops in the NIC value can localize the interference source with good accuracy compared to the original jamming source, justifying a high correlation between NIC value change and localizing the jamming source location. Furthermore, the percentage increase in data added from interpolation suggests that predicting gaps (especially during potential interference) adds a significant amount of missing flight path information, which can be useful in other IDL algorithms.
Published in:	Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022) September 19 - 23, 2022 Hyatt Regency Denver Denver, Colorado
Pages:	947 - 957
Cite this article:	Dacus, Michael, Liu, Zixi, Lo, Sherman, Walter, Todd, "Improved RFI Localization Through Aircraft Position Estimation During Losses in ADS-B Reception," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 947-957. https://doi.org/10.33012/2022.18529
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In