Suppressing Mainlobe Deceptive Jammers via Two-Low-Rank Matrix Decomposition in FDA-MIMO Radar

Frequency diverse array multiple-input multiple-output radar is capable of suppressing mainlobe deceptive jammers owing to its extra degrees-of-freedom in the range domain. However, it fails to suppress the rapidly generated mainlobe jammers, which occur within the same transmit pulse as the true target. To address this challenge, an innovative approach that integrates the two-low-rank matrix decomposition (2-LRMD) model is devised. At the modeling stage, leveraging the low-rank characteristics in the spatial domain, the data matrix is decomposed into two low-rank matrices, respectively, representing the true target and mainlobe deceptive jammers. This decomposition is then formulated into an optimization problem, where alternating direction method of multipliers (ADMM) is utilized to isolate the target and jammer signals. Furthermore, to eradicate residual jammer components within the received target signal, truncated singular value decomposition is adopted. On the other hand, the convergence properties of the proposed 2-LRMD-ADMM algorithm are rigorously studied, and its computational complexity is analyzed. Numerical results showcase the effectiveness of our solution especially in suppressing rapidly generated mainlobe jammers, where comparisons among different frameworks and methodologies are carried out.