Composite Indicator for Detecting and Localizing Time-Varying RFI in SAR Raw Data

Radio frequency interference (RFI) seriously affects the processing of synthetic aperture radar (SAR) data and the interpretation of images. The accuracy of RFI detection, which is the first step of RFI suppression, directly affects the follow-up work. However, the implementations of existing RFI detection methods mostly rely on the assumption that the flatness of the range spectrum will be destroyed by strong RFI, making these methods ineffective in the face of weak energy RFI. Furthermore, in these methods, the problem of accurately localizing the RFI is rarely considered, resulting in a considerable loss of imaging signals during subsequent RFI suppression. In this manuscript, a method of time-varying RFI detection and localization is proposed with two steps that are based on the flatness and symmetry of the SAR range spectrum. This method can effectively avoid missed detection and accurately find the location of RFI with weaker energy in an echo. The proposed method first detects the existence of RFI along the azimuth direction. Moreover, the proposed composite index is used. The position of the RFI is then found along the range direction, and the log ratio operator and eigenvalue decomposition (EVD) are used. Experimental results on both simulated and raw data obtained by airborne and spaceborne SAR validate the advantages of the suggested method in detecting and locating weak energy RFI.