A SAR RFI Suppression Method Based on Low-Rank Guided CUR Decomposition in Complex Electromagnetic Environment

Synthetic Aperture Radar (SAR) technology is extensively utilized in complex electromagnetic environments. Radio Frequency Interference (RFI) has emerged as a critical factor degrading the quality of SAR images, target recognition, and mapping accuracy. The strong time-frequency correlation between signals and interference poses significant challenges for traditional RFI suppression methods, making it difficult to effectively distinguish between them. This paper introduces an RFI suppression method based on low-rank guided CUR decomposition. To address the strong time-frequency correlation and echo aliasing characteristics inherent in SAR data, this method first accurately extracts core features of the interference using Singular Value Decomposition (SVD)-guided row and column selection. Secondly, within the alternating projection framework, low-rank guided sampling leverages the inherent low-rank structure and interference characteristics of SAR data to dynamically select the most representative rows and columns for CUR decomposition. This approach not only preserves the low-rank structure of the data but also significantly mitigates the loss of useful signals, thereby enhancing the suppression effect and demonstrating substantial potential for practical application.