GNSS Signal Disturbance Detection and Classification Based on LEO Satellite Measurements

Kahn-Bao Wu, Y. Jade Morton, S. Tim Dittmann, H. Chang

Peer Reviewed

Abstract:	This paper presents a radial basis function (RBF) support vector machine (SVM)-based machine learning (ML) model to detect and classify disturbances observed in dual-frequency carrier phase measurements obtained by side-looking antennas on low Earth orbiting (LEO) CubeSats. The model is designed to distinguish radio frequency interference, satellite oscillator anomaly, and ionospheric disturbances. The trained ML model is applied to measurements collected by Spire Global CubeSats in 2021. The testing and validation results show that the RBF SVM achieved an average ???? score of 90%. The results show that RFI occurrence is 8.0% and the ionosphere disturbance occurrence is 0.3% in the data sets. Other possible causes of the signal disturbances such as, satellite attitude variations, satellite communications with ground monitoring stations, and low SNR associated with signals entering low antenna gain areas are all group into the RFI class. Future work will further distinguish these sources.
Published in:	Proceedings of the 2024 International Technical Meeting of The Institute of Navigation January 23 - 25, 2024 Hyatt Regency Long Beach Long Beach, California
Pages:	417 - 425
Cite this article:	Wu, Kahn-Bao, Morton, Y. Jade, Dittmann, S. Tim, Chang, H., "GNSS Signal Disturbance Detection and Classification Based on LEO Satellite Measurements," Proceedings of the 2024 International Technical Meeting of The Institute of Navigation, Long Beach, California, January 2024, pp. 417-425. https://doi.org/10.33012/2024.19481
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In