A method for suppressing SAR radio frequency interference based on deep learning in time-frequency domain

Synthetic-aperture radar is widely deployed due to its high resolution, such as deployed in battlefield monitoring, geographic mapping, disaster assessment and other fields. Radio frequency interference, as an active interference method, mainly emits electromagnetic waves to make the radar receiver inevitably receive interference signals while receiving echoes, forming a masking effect on the images, thereby reducing the quality of SAR images. Many filtering algorithms have been proposed to address the aforementioned issue, however, these algorithms generally lack generalization and the real-time performance is insufficient. Therefore, in order to solve the above problems, this paper proposes a deep learning filtering algorithm and constructs a simulation dataset in the time-frequency domain. The final experimental results show that the PSNR of the time-frequency map is improved to 35.4dB by our method, and the interference on the SAR image is effectively filtered out. What's more, this method has good generalization and can work well against various RFI styles.