A Radio Frequency Interference Detection and Analytical Geolocation Method Demonstrated from LEO with an End-To-End Simulator and CYGNSS Raw Data

Radio frequency interference (RFI) presents a growing threat to applications, ranging from surveying and construction to remote sensing and GNSS reflectometry (GNSS-R). RFI inevitably degrades the quality of GNSS-R observation data and data inversion. Due to ongoing geopolitical conflicts, the need to detect and geolocate these RFI sources worldwide has been increasing. This paper presents an analytical method based on the Doppler shift and Doppler rate measurements using a single Low Earth Orbit (LEO) satellite to detect and geolocate RFI transmitters. The methods are demonstrated using RFI-simulated data generated with the Orolia Skydel GNSS simulator and validated against historical Cyclone GNSS (CYGNSS) raw IF data. The geolocation method provides an unambiguous, converged, and stable solution through an optimisation process that estimates the location of the RFI transmitter. The RFI detection method is based on the short-time Fourier transform (STFT), which is robust in low signal-to-noise ratio (SNR) environments, operating reliably down to -35 dB. The geolocation method is computationally efficient, operating on short data windows (30-60 seconds) and accommodating different types of interference. Two CYGNSS raw IF datasets were selected for detection and geolocation. The geolocation algorithm produced a converged and consistent estimated RFI transmitters at the global minimum of the cost function, while also providing accurate vertical positioning of the RFI sources through the use of the Copernicus GLO-30 Digital Elevation Model (DEM). The geolocation accuracy for the simulated RFI scenarios was approximately within (20-50) meters, while for overhead scenarios it was 800 meters, reflecting the sensitivity of the analytical model that has been addressed in this paper. The end-to-end simulation allows assessments of the accuracy in a variety of scenarios, while the CYGNSS raw data demonstrate the method and techniques applied to current in-orbit data.