论文部分内容阅读
基于压缩感知的下视三维合成孔径雷达(DL 3DSAR)成像算法可以采用低于Nyquist采样率的采样数据实现平台正下方的高分辨成像。但是已有的算法在跨航向重构时采用的大都是单测量向量(SMV)模型,存在重构耗时长、受噪声干扰大的缺点。从单测量向量的推广形式即多重测量向量(MMV)模型出发,将DL 3DSAR中的跨航向处理与沿航迹向处理顺序交换,利用多重测量向量恢复具有相同稀疏结构的跨航向信号,提出了一种基于MMV模型的DL 3DSAR成像算法。相比于SMV模型的DL 3DSAR成像算法,该算法在运算时间、抗噪性能及重构精度方面均有所提高,并通过仿真实验验证了算法的有效性。
The DL 3DSAR imaging algorithm based on compressive sensing can realize high-resolution imaging right below the platform using sampled data below the Nyquist sampling rate. However, most of the existing algorithms adopt single-measurement vector (SMV) model in cross-heading reconstruction, which has the disadvantage of long reconstruction time and large noise interference. Based on the multi-measurement vector (MMV) model, which is the generalized form of single measurement vector, the cross-heading process in DL 3DSAR is exchanged with the process sequence along the flight path, and the multi-measurement vector is used to recover the cross-heading signal with the same sparse structure. A DL 3DSAR Imaging Algorithm Based on MMV Model. Compared with the DL 3DSAR imaging algorithm based on SMV model, this algorithm has improved the computing time, anti-noise performance and reconstruction accuracy. The simulation results show the effectiveness of the algorithm.