Characteristics of RFI Determined From Kurtosis Using the SMAP Radiometer

Radio frequency interference (RFI) is a problem in microwave remote sensing even for sensors operating in the protected band at 1.4 GHz (L-band). Unfortunately, little is known about the sources of the interference, which complicates the design of systems to deal with it. A unique feature of the Soi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on geoscience and remote sensing 2024, Vol.62, p.1-11
Hauptverfasser: Le Vine, David M., De Matthaeis, Paolo
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Radio frequency interference (RFI) is a problem in microwave remote sensing even for sensors operating in the protected band at 1.4 GHz (L-band). Unfortunately, little is known about the sources of the interference, which complicates the design of systems to deal with it. A unique feature of the Soil Moisture Active Passive (SMAP) radiometer is that it comprises an array of tools to detect RFI, including spectral information and kurtosis in addition to the more conventional time-domain thresholding. This article reports the results of an investigation to determine if the kurtosis ( K ) can be combined with the information about the spectrum of the RFI to identify characteristics of the source of the RFI. The SMAP conical scan permits accurate location of the source of RFI resulting in a few cases where identification of the source of RFI was possible. The known sources include both radar and domestic electronics. Evidence is presented that RFI with K < 3 is most likely to be continuous and confined to a narrow frequency (e.g., faulty electronics) and that RFI with K > 3 and large is likely to be associated with short pulses (e.g., radar, but there are exceptions). A critical parameter in determining the magnitude of the kurtosis is the duration of the RFI relative to the integration time of the radiometer.
ISSN:0196-2892
1558-0644
DOI:10.1109/TGRS.2023.3338964