Grid-Matched Sparse Recovery STAP Framework via Spectrum Diagonal Linearization for BiSAR Clutter Suppression

Clutter suppression is an effective technology to achieve high value time-sensitive ground moving target indication. However, for bistatic synthetic aperture radar (BiSAR), clutter exhibits characteristics such as clutter spectrum expansion and nonstationarity, which poses a huge challenge to traditional space-time adaptive processing (STAP) methods. Currently, the booming sparse recovery based STAP (SR-STAP) methods can achieve high precision space-time spectrum reconstruction with few samples, greatly avoiding the clutter nonstationarity. However, the inherent atomic off-grid problem for SR-STAP methods can also induce clutter space-time spectrum expansion, resulting in a serious deterioration of clutter suppression performance. To address the aforementioned issues, a grid-matched SR-STAP framework via space-time spectrum diagonal linearization (SDL) for BiSAR clutter suppression is proposed in this paper. Its innovation lies in the construction of the SDL matrix, which achieves a diagonalized linearization of the space-time spectrum in BiSAR. This approach effectively solves the atomic off-grid problem arising from the high-order nonlinear coupling characteristics of BiSAR clutter’s space-time distribution. The proposed framework mainly consists of three steps. Firstly, to achieve consistent compensation of range cell migration and Doppler frequency rate, the joint Keystone transformation and nonlinear chirp scaling processing is applied. Secondly, to convert the BiSAR space-time spectrum into that with diagonalized linearization, the SDL matrix is constructed with prior information. Finally, to obtain the high precision grid-matched BiSAR space-time spectrum reconstruction result, the inverse SDL transformation is derived, then the matched space-time filter is designed. The simulation and experimental results both verify the effectiveness of newly proposed framework.