In complex electromagnetic environments, synthetic aperture radar (SAR) is severely affected by radio frequency interference (RFI) from other systems, such as ground-based radar, cellular networks, and global positioning systems, and this interference cannot be neglected. Pulse RFI (PRFI), a common form of RFI, can hinder SAR signal processing and image interpretation to varying degrees. The time-domain notch filtering method designed for mitigating PRFI can locate and mitigate the evident PRFI covered in SAR echo data, but it is helpless against PRFI hidden in a strong echo signal. In this article, a three-step approach is proposed to tackle the PRFI problem. In the proposed approach, the first step is to detect and locate PRFI; this is based on eigenvalue decomposition (EVD) and the short-time Fourier transform (STFT). The second step is to notch PRFI, and this is based on a time-domain notch filter. The third step is to recover the notched signal using a novel matrix completion strategy, which integrates with a robust low-rank matrix completion (LRMC) technique—i.e.,the singular value thresholding (SVT) algorithm—and a well-known Lagrange interpolation technique. Experimental results via simulated SAR data, Sentinel-1 level-0 raw data, and L-band airborne SAR raw data demonstrate the performance of the proposed approach.

In synthetic aperture radar (SAR) data processing, radio frequency interference (RFI) has been recognized as a challenging issue, which can significantly degrade the image quality, including amplitude, phase, geometry, and so on. Most of the RFI sources are direct waves transmitted from the ground. Recently, a new form of RFI, namely, the mutual RFI (MRFI), which is a kind of scatter-wave RFI originating from the nearby ground area simultaneously illuminated by different SAR satellites, has been reported. In this article, the signatures of MRFI are characterized, and an example case study between Chinese GaoFen-3 (GF-3) and European Sentinel-1A (S-1A) is analyzed. In order to remove the artifacts caused by MRFI on SAR images, a novel RFI detector based on spectrum energy cancellation (SEC), which has the capability of detecting MRFI, is developed. Two methods, i.e., the traditional notch filtering method and an improved eigen-subspace projection (ESP) method proposed in this article, are employed to mitigate MRFI. The former method has robust mitigation performance with a fast processing speed, whereas the latter has an improved mitigation accuracy and lower sidelobes for strong scatterers. The methods are designed for cases in which the MRFI has a different radar center frequency than the useful signal, and some of the data are free from MRFI. The conditions required for the proper use of the approach are discussed. In contrast to the conventional mitigating methods performed on the raw data domain, the proposed approach begins with the focused single-look complex (SLC) SAR data. The effectiveness of the proposed approach is demonstrated on several GF-3 SLC SAR images, which were contaminated by MRFI from the S-1A.

Synthetic aperture radar (SAR) often shares spectrum with other systems, such as radio, TV, cellular network, and so on, which are likely to produce radio frequency interference (RFI). Pulse RFI, which hinders SAR signal processing and image interpretation severely, is a common form of RFI and cannot be neglected. The simple and easy-to-implement frequency-domain notch filtering (FNF) method has been widely used to mitigate narrowband pulse RFIs. However, a well-known problem with abnormal sidelobe effect, which is caused by missing spectrum gaps due to the notch operation, is aroused when using FNF. In this letter, a novel time-domain notch filtering (TNF) is proposed. In the proposed method, pulse RFI occurrences are detected and notched by a simple log-ratio operator in a pulse by pulse manner. Then, missing-data iterative adaptive approach (MIAA) is performed to recover the notched signal to avoid ghosts. Experimental results via simulated and real L-band airborne SAR raw data validate the performance of the proposed method.

Currently, the spaceborne synthetic aperture radar (SAR) system transmits a great deal of data to the ground processing station and generates massive images daily, only a tiny fraction of which contains radio frequency interference (RFI). However, most of the existing RFI detection methods are based on the prior conditions in which the known image contains interference. In fact, it is difficult to learn whether SAR images contain RFIs without prescreening, so it is of great significance to the rapid and real-time screening and detection of RFI in SAR images. This article proposes a method to screen and detect RFI from massive SAR images simultaneously. 1) We construct an approximate RFI-free background image by using the preprocessed time-series SAR images acquired in the past. 2) We generate difference images based on the image change detection method and analyze them by using an adaptive threshold, then calculate the entropy of all difference images to complete the preliminary screening of the RFI-containing). According to the preliminary results, we remove the RFI-containing parts; after reconstructing the background, we repeat step 2 with the images to be detected and obtain the final screening and detection results. Massive experimental results based on Sentinel-1 images validate the performance of the proposed method.

Benefiting from the rapid development of deep learning, the field of synthetic aperture radar (SAR) ship detection has been promoted with renewed vigor. However, due to the unique characteristics of SAR ship detection, several challenges have been encountered in the process of fusing SAR ship detection. On the one hand, some special properties of SAR images, including low resolution, small targets, dense ship inshore arrangements, and background clutter noise, increase the difficulty of detection; on the other hand, satellite detection requires a high degree of real-time and lightweight, and most traditional models fail to satisfy the requirements in two areas. After considering the above issues, we combine the advantages of space-to-depth convolution (SPDConv) and inception depthwise convolution (InceptionDWConv) to design SPD-InceptionDWConv (SIDConv), which resolves the above problems with fewer parameters and operations. Moreover, based on SIDConv, we also design light SIDConv (LSIDConv) and efficient SIDConv (ESIDConv) to further reduce the model complexity and to accelerate the detection process. With the above module as the core and YOLOv5 as the base, we propose a lighter and faster detector named LFer-Net, which has only 0.8M parameters and 1.9G FLOPs. Extensive experimental results on the SSDD and HRSID datasets confirm that our model is not only superior to other models with 98.2 and 90.6 accuracy but also achieves detection speeds of 144 and 96 frames per second (FPS), respectively.

Synthetic aperture radar (SAR), an active remote sensing equipment, shares the spectrum with devices in the same frequency band and is therefore easy to affect by pulse radio frequency interference (PRFI). The wideband version of PRFI, i.e., linear-frequency-modulation PRFI (LFM-PRFI), derived from the ground- and space-based radar sensors is a challenging issue for SAR because it usually has a large bandwidth and pulse width compared with the traditional PRFI. The well-known notch filtering methods, including time- and frequency-domain versions, are robust approaches against PRFI, which has been integrated into some SAR ground processing systems. However, the down-range frequency domain notch filtering will increase the sidelobes of the targets when mitigating wideband LFM-PRFI, whereas the time-domain version will introduce ghosts when notching large pulse-width ones. This paper proposed a filtering method that consists of three steps to tackle the above problems. The first step is focusing the energy of LFM-PRFI down-range and cross-range simultaneously, which can be done by match filtering and Fourier transform. The second step is mitigating the LFM-PRFI, which can be done by a designed filter. The last step is data restoration, which can be done by multiplying the conjugate of the phase function used before. Numerical experiments based on simulated SAR data and measured spaceborne SAR data acquired by European Sentinel-1 are performed to test the mitigation performance, which verified the effectiveness and superiority of the proposed approach.