3-D High-Resolution Imaging and Array Calibration of Ground-Based Millimeter-Wave MIMO Radar

Abstract

Ground-based (GB) radar imaging can provide mapping of interesting terrains and accurately monitor deformation in a noncontact manner, which has found a wide application in landslide warning and so on. In this article, a novel 3-D synthetic aperture radar (SAR) imaging and array phase error calibration algorithm is proposed using the micro multiple-input–multiple-output (MIMO) millimeter-wave (mmW) radar sensor. The framework of Doppler-division multiplexing (DDM) MIMO-SAR is designed to be suitable for a moving platform, i.e., a rail-mounted-based radar system. Then, a novel 3-D SAR imaging algorithm is presented to integrate the backprojection (BP) 2-D imaging and super-resolution MIMO array imaging by using an atomic norm minimization (ANM) method. Benefiting from the off-grid characteristic of atomic norm, the proposed ANM 3-D SAR imaging algorithm can effectively overcome the drawback of discrete errors compared with conventional methods using the discrete dictionary of Fourier transform. Meanwhile, a minimum entropy algorithm is proposed to accurately estimate and correct the MIMO array phase errors, which is classified as the self-calibration method. Finally, the numerical experiments using the measured data from a 77-GHz $4$ $\times $ $8$ radar are performed to confirm the proposed GB 3-D SAR imaging algorithm.