Super Resolution DOA Based on Virtual Sparse Extension Array for FMCW Automotive Radar

Automotive radar has been emerging as a key technology enabling intelligent and autonomous features in modern vehicles. However, its great weakness is the poor angular resolution compared with its competitors, such as LiDAR or camera. Thus, one of the most challenging problems automotive radars faced is how to achieve high angular resolution, especially with the restriction of limited array aperture and in high dynamic environment. In this paper, a novel approach for high angular resolution DOA is proposed to separate two targets in the same range and Doppler cell by introducing one completely new concept of virtual sparse extension array. Based on the radar returns from the physical array, an extremely large virtual sparse aperture can be formed in a unique way. The simulation results demonstrate that, for a typical time division modulation (TDM) multiple-input multiple-output (MIMO) automotive radar with 3 transmitters and 4 receivers, the angular resolution can reach 0.01°, which is far better than other algorithms. In addition, the proposed method only needs single snapshot, and this is especially valuable to deal with the practical high dynamic environment.