An Efficient Sparse Representation Method for Passive Radar

Passive radar (PR) commonly estimates target parameters by calculating the cross-ambiguity function (CAF), which is prone to generating a wider main lobe and higher sidelobes, leading to issues such as weak targets being masked and adjacent targets being difficult to distinguish. A parameter estimation method for PR based on sparse representation (SR) is proposed to address the above challenges. First, an SR model based on signal segmentation and Fourier transform is proposed to address the issue of excessively large dictionary matrix (DM) by using the fast Fourier transform (FFT). Then, an orthogonal matching pursuit (OMP) algorithm based on detection threshold (DT-OMP) is proposed to adaptively determine the number of atoms to be selected by a preset threshold. Furthermore, a model mismatch correction method for SR (MMC-SR) is proposed to achieve accurate estimation of target parameters in off-grid situations. Simulations and practical experiments have shown that the proposed method can effectively mitigate the influence of wider main lobe and higher sidelobes of CAF, thereby improving resolution and providing a refined estimation of target parameters, showcasing significant practical application value.