A Weighted Low-Rank and Sparse Constraint-Based Multichannel Radar Forward-Looking Imaging Method

Radar super-resolution imaging methods with joint low-rank and sparse constraints have garnered increasing attention. However, in complex imaging scenarios, the low-rank property of the signal matrix is often not prominent, which limits the performance of directly applying low-rank constraints in super-resolution imaging. To address this issue, this article proposes a weighted low-rank and sparse constraint-based multichannel radar forward-looking super-resolution imaging method. First, the proposed method calculates weighting coefficients using the covariance matrix of radar echoes and applies weighted constraints to the signal matrix, thereby enhancing its low-rank property and significantly improving forward-looking super-resolution imaging performance. Then, in solving the optimization problem, the alternating direction method of multipliers (ADMM) is employed to decompose variables and reduce the complexity of the solution. To further enhance computational efficiency, the symmetry of the weighting matrix and the characteristics of the dictionary matrix in sparse imaging, specifically as a partial Fourier dictionary, are leveraged. A fast matrix inversion method based on eigenvalue decomposition is proposed to mitigate the computational complexity induced by super-resolution requirements. Finally, the effectiveness and superiority of the proposed method in complex scenarios are validated through comparative experiments on simulated and measured data.