Fast Forward-Backward Hankel Matrix Completion for Automotive Radar DOA Estimation Using Sparse Linear Arrays

Automotive multiple-input multiple-output (MIMO) radar with sparse linear arrays is a cost-effective solution to achieve large aperture size with low hardware cost and reduced mutual coupling. The challenges associated with automotive MIMO sparse linear arrays are the high sidelobes, which might result in angular detection errors. This paper presents a fast forward-backward Hankel matrix completion and matrix pencil method for joint array interpolation and super-resolution single-snapshot angle finding by exploiting the structure of Hankel matrix. The novelty of the proposed approach lies in two parts. It not only saves the computational cost of singular value decomposition (SVD) in each iteration of the matrix completion, but also increases the degrees of freedom to construct a low-rank matrix with larger dimensions using the same number antenna elements, as a result of which, more targets can be completed and estimated with better accuracy. Numerical results demonstrate the effectiveness and efficiency of the proposed method.