Research on Methods for Narrow-Band Interference Suppression in Synthetic Aperture Radar Data

Abstract

Narrow-band interference (NBI) poses a hindrance to high-quality imaging for synthetic aperture radar (SAR). It is an under-determined single-channel separation problem. In this paper, we addressed the NBI suppression problem by introducing two advanced data-driven nonparametric techniques: 1) the eigen subspace filtering and 2) the independent component analysis (ICA). Both of these two methods utilize the statistical difference between the useful radar echoes and NBI. The interference-contaminated pulse is decomposed into a set of basis signals, from which the bases corresponding to NBI are selected out. Then, the contribution of NBI is excised by filtering out the corresponding interference components. The performances of these advanced methods are compared with the conventional notch filtering method. The experimental results on real datasets show the effectiveness of the proposed methods.