High-Resolution Point-Cloud Imaging With Doppler Division MIMO Radar Based on the 2-D Hybrid Sparse Array

Abstract

Automotive radar point-cloud imaging plays an important role in advanced driver assistant systems (ADASs), and most vehicle-mounted radars improve the angular resolution by the time-division multiplexing multiple-input and multiple-output (TDM-MIMO). However, the performance of TDM-MIMO radar suffers seriously from the transmitted energy loss, serious Doppler ambiguity, and the coupling phase induced by the switching delay. In this article, we have proposed a 4-D point-cloud imaging method based on the Doppler division multiplier access (DDMA) MIMO radar and have used the sparse array to balance the contradiction between the Doppler ambiguity and the angle resolution. First, a 2-D hybrid sparse array, both the transmitted array and the received array being sparse linear array (SLA), is designed to mitigate the Doppler ambiguity to a certain extent. Sequentially, targets’ locations in space are been focused by taking advantage of the low rankness of the snapshot matrix, and accordingly, facing the problem of decreased signal-to-noise ratio (SNR) directly by the hybrid sparse snapshot matrix, we have proposed jointly low rankness and sparsity based on the matrix factorization (JLSMF) algorithm to obtain the uniform snapshot matrix and the sparse locations of scattering points. Compared with previous achievements, the proposed algorithm has a better performance, lower computation complexity, smaller recovery error, and so on. Finally, simulation experiments have validated the effectiveness of the proposed algorithm. Besides, the proposed algorithm has great reference value in other fields, such as inverse synthetic aperture radar (ISAR), magnetic resonance imaging, and so on.