Enhanced Imaging for Forward Looking MIMO SAR Via Un-Supervised Deep Basis Pursuit

Nowadays, radar image reconstruction is becoming important in the context of advanced driver assistance systems especially for all weather conditions. In this paper, we present image reconstruction with deep learning based methods on forward looking multiple-input multiple-output array synthetic aperture radar (FL-MIMO SAR). We present deep basis pursuit (DBP) method to solve for convolutional neural network (CNN) weights with unsupervised learning (i.e. without ground-truth) and present modified back projection (MBP) algorithm to reconstruct SAR image with enhanced angular resolution. We present experimental results to verify our proposed methodology on both simulation and real data.