LAUSR

Radio frequency interference (RFI) seriously deteriorates the quality of the retrieval of geophysical parameters, e.g., Earth surface moisture and ocean salinity, measured in microwave interferometric radiometry (MIR). The accurate detection of RFI sources is crucial for locating these illegal sources and mitigating their impact. In this article, we propose a new method based on reweighted $\ell _{1}$ -norm minimization to detect RFI sources. First, we exploit the sparsity of RFI sources in the spatial domain and formulate the RFI detection as a problem of reweighted $\ell _{1}$ -norm minimization, by which the RFI signals can be well recovered and the background noises can be suppressed. Then, we present two algorithms, termed RL1 and NRL1, to achieve RFI source detection. The RL1 algorithm employs a fast iterative shrinkage thresholding (FIST) technique, and the NRL1 algorithm combines the FIST with a neighbor-reweighting strategy that helps to further enhance the RFI target regions. Finally, simulations and experiments using Soil Moisture and Ocean Salinity (SMOS) satellite data demonstrate the superiority of the proposed method on the RFI-signal-to-background ratio (RSBR) in recovered images and the detection performance of RFI sources, compared with the existing RFI processing methods in MIR.