A Conditional Denoising Diffusion Probabilistic Model for RFI Mitigation in Synthetic Aperture Interferometric Radiometer

Abstract

In Earth remote sensing, spatial-frequency domain visibility samples are inversely transformed into spatial-domain brightness temperature (BT) images through the signal processing pipeline of synthetic aperture interferometric radiometers (SAIR). However, L-band radio-frequency interference (RFI) contaminates the measured visibilities and severely degrades BT image quality, thereby impairing geophysical parameter retrieval. To address this issue, we propose VFDM, a Visibility-Function Diffusion Model based on Denoising Diffusion Probabilistic Models (DDPM), to mitigate RFI in the spatial-frequency domain while preserving fine-scale structures consistent with natural scene statistics. Furthermore, we construct a comprehensive dataset comprising more than ten thousand pairs of RFI-free natural scene visibility sample sets and their corresponding simulated contaminated counterparts, categorized by varying RFI intensities, numbers, and distributions. Finally, comprehensive experiments on both simulated and real-world data demonstrate the effectiveness and robustness of the proposed VFDM-based approach.

Figures (3)

• Figure 1: Overview of the Proposed VFDM. The diffusion process starts from $\mathbf{R}_{S,0}$ and Gaussian noise is gradually added to $\mathbf{R}_{S,t-1}$ to obtain $\mathbf{R}_{S,t}$ until $t = T$. The reverse process starts from random noise and generates $\mathbf{R}_{S,t-1}$ from $\mathbf{R}_{S,t}$. The contaminated covariance matrix is incorporated into the reverse process as a condition to guide the reconstruction of the natural covariance matrix.
• Figure 2: RFI mitigation examples on simulated data. (a) Scene 1 (snapshot ID: 691841314); (b) weak RFI contamination of scene 1 and (c)-(f) corresponding results after applying CLEAN, RPCA, RNN-DFT, VFDM respectively. (g) Scene 2 (snapshot ID: 691834271); (h) hybrid RFI contamination of scene 2 and (i)-(l) corresponding results after applying CLEAN, RPCA, RNN-DFT, VFDM respectively.
• Figure 3: RFI mitigation examples on real data. (a) Scene 3 (snapshot ID: 691840761) and (b)-(e) corresponding results after applying CLEAN, RPCA, RNN-DFT, VFDM respectively. (f) Scene 4 (snapshot ID: 691840910) and (g)-(j) corresponding results after applying CLEAN, RPCA, RNN-DFT, VFDM respectively.