TDM-MIMO Automotive Radar Point-Cloud Detection Based on the 2-D Hybrid Sparse Antenna Array

Automotive radar plays an important role in the field of advanced driver assistant systems (ADASs), in the detection of unmanned aerial vehicles (UAV), and so on. Most automotive radar products can only estimate the azimuth, and however, it is necessary for automatic driving technology to acquire height information. In this article, we propose a signal processing algorithm based on the 2-D sparse multiple-input–multiple-output (MIMO) antenna array, which can detect range, velocity, azimuth, and elevation of scattering points, and this location information can form a clear point-cloud image. First, a 2-D hybrid sparse array is designed with the advantage of reducing hardware resources, such as the number of transceiver channels. Based on the sparse antenna array, a robust matrix completion recovery algorithm, combining the low-rankness of the snapshot and the sparsity of the grid map, is further proposed to extract azimuth and elevation, and alleviate the impact of grating lobes on them. Sequentially, azimuth and elevation of targets can be shown through the correction of motion-induced phase in time-division MIMO (time-division multiplexing (TDM)-MIMO) radar and by the 2-D fast Fourier transform (2DFFT) digital beamforming (DBF). Finally, many experiments validate that a point-cloud image, such as LiDAR, can be shown, which will be hopefully applied for ADASs, UAVs, and other fields.