Radio Frequency Interference Detection and Localization in Sentinel-1 Images
Xiangguang Leng, Kefeng Ji, Gangyao Kuang
Abstract
The C-band Sentinel-1 synthetic aperture radar (SAR) images are affected by radio frequency interference (RFI), and this article proposes an RFI detection and localization method for these images. First, a detailed analysis of RFI based on the generation mechanism is provided, which shows that RFI in the ground range detected (GRD) products differs from common backscattering not only in frequency spectrums but also in radiometric characteristics. Then, an RFI index (RFII) is proposed for RFI detection, which takes full advantage of unique RFI characteristics in dual-polarization GRD images. Finally, both detection results in ascending and descending passes are used to locate the ground RFI sources, assuming that the presence of RFI is persistent. Theoretical analyses show that the location area is a diamond of approximately 88.76 km <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . Experimental results demonstrate that the proposed method performs quite well on Sentinel-1 GRD data. It provides a convenient and tractable tool for assessing Sentinel-1 data quality as well as for monitoring RFI. Thus, Sentinel-1 measurements can be used to monitor C-band RFI, which is an important task of electromagnetic spectrum management.