Machine Learning Detection and Classification of Internal Radio Frequency Interference in Spire Global GNSS-RO Signals

Jason Li, Hyeyeon Chang, Tim Dittmann, and Jade Morton

Peer Reviewed

Abstract:	GNSS radio occultations received from commercial low Earth orbit satellites offer a cost-effective way to profile Earth’s atmosphere and ionosphere. However, these systems are susceptible to disturbances from non-geophysical sources, such as internal satellite communications and external GNSS jammers, and their effects can often exceed heuristic thresholds set to detect natural ionospheric scintillations. This study focuses on distinguishing ionospheric disturbances from both external and internal RFI in Spire Global’s GNSS-RO data with the development of a convolutional neural network trained with labeled dual frequency, 50 Hz GNSS-RO SNR and excess phase observations. Index Terms—GNSS-RO, Interference, Jamming, Machine Learning
Published in:	2025 IEEE/ION Position, Location and Navigation Symposium (PLANS) April 28 - 1, 2025 Salt Lake Marriott Downtown at City Creek Salt Lake City, UT
Pages:	288 - 293
Cite this article:	Li, Jason, Chang, Hyeyeon, Dittmann, Tim, Morton, Jade, "Machine Learning Detection and Classification of Internal Radio Frequency Interference in Spire Global GNSS-RO Signals," 2025 IEEE/ION Position, Location and Navigation Symposium (PLANS), Salt Lake City, UT, April 2025, pp. 288-293.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In