Suppression of Compound Mainlobe Interferences Exploiting Nonconvex Low-Rank and Sparse Decomposition in FDA-MIMO Radar

This paper addresses the issue of suppressing com pound mainlobe interferences in frequency diverse array multiple input multiple-output (FDA-MIMO) radar, including the rapidly generated jammers as well as burst interference, based on noncon vex low-rank and sparse decomposition (NLRSD). In particular, the problem is formulated by leveraging the low-rank characteristics of deceptive jammers and the true target, as well as sparse char acteristic of burst interference in the spatial-temporal domain. To address the problem, a NLRSD method is developed, exploiting the unitary invariance of the γ norm and the Moreau-Yosida operator to reformulate the problem as a singular value optimization task. The alternating direction method of multipliers (ADMM) is implemented to separate the target and interference signals. Furthermore, the convergence performance, which satisfies the Karush-Kuhn-Tucker (KKT) conditions, as well as the computational complexity of the proposed NLRSD-ADMM algorithm are rigorously analyzed. Numerical results and comparisons with existing frameworks and methods validate the effectiveness of the proposed NLRSD-ADMM method in suppressing both rapidly generated mainlobe jammers and burst interference.