Bistatic Aperture Completion via Untrained Fourier Convolutional Networks

Effective techniques for solving the inverse scattering problem is crucial for many areas of electromagnetic applications, including distributed radar, ground penetrating radar, and radio-frequency medical imaging. It is well established that the inverse scattering problem is particularly challenging due to inherent non-linearity and ill-posedness. As such performance of many algorithms is highly dependent on a high degree of space-frequency diversity.