Near-field Sparse Array mmWave Radar Imaging by Matrix Completion with Improved SVT Algorithm

A Near-field sparse array millimeter wave (mmWave) radar imaging method is proposed using matrix completion via an improved singular value thresholding (SVT) algorithm. Different from traditional SVT algorithm used for matrix completion, the improved SVT algorithm remarkably reduces the number of iterations through optimization of the iteration step size. Feasibility of the proposed method is experimentally verified by using a 77-GH TI mmWave radar to collect the full sampling data and sparse sampling data respectively. The proposed method is a good solution to reducing the computational complexity while maintaining high-quality imaging capabilities.