系统工程与电子技术 ›› 2019, Vol. 41 ›› Issue (2): 304-310.doi: 10.3969/j.issn.1001-506X.2019.02.12

• 传感器与信号处理 • 上一篇    下一篇

基于l1正则化的多通道滑动聚束SAR成像

徐志林1,2,3, 魏中浩1,2,3, 吴辰阳1,2,3, 张冰尘1,2   

  1. 1. 中国科学院电子学研究所, 北京 100190;  2. 空间信息处理与应用系统技术重点实验室,北京 100190;  3. 中国科学院大学, 北京 100190
  • 出版日期:2019-01-25 发布日期:2019-01-25

Multichannel sliding spotlight SAR imaging based on l1 regularization

XU Zhilin1,2,3,WEI Zhonghao1,2,3, WU Chenyang1,2,3, ZHANG Bingchen1,2   

  1. 1. Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, China;
    2. Key Laboratory of Technology in Geospatial Information Processing and Application System, Beijing 100190, China;
    3. University of Chinese Academy of Sciences, Beijing 100190, China
  • Online:2019-01-25 Published:2019-01-25

摘要: 提出了基于l1正则化的多通道滑动聚束合成孔径雷达(synthetic aperture radar, SAR)稀疏成像算法。该方法将偏置相位中心天线(displaced phase center antenna,DPCA)技术和l1正则化策略结合来解决非均匀采样带来的方位模糊问题,并利用方位距离解耦算法降低计算复杂度。当非均匀度较大时,所提方法相比于基于多普勒频谱重建的匹〖JP3〗配滤波器组方法能更有效抑制方位模糊,具有更大的距离向测绘带宽潜力。该方法能有效抑制噪声和旁瓣,提高目标背景比,从而提高成像性能。通过仿真和实际数据实验,验证了所提算法的有效性。

Abstract: l1 regularization based multichannel sliding spotlight synthetic aperture radar (SAR) imaging method is proposed. The proposed method combines the displaced phase center antenna (DPCA) technology with the l1 regularization scheme to solve the azimuth ambiguities problem caused by nonuniform sampling, and introduces azimuthrange decouple algorithm to reduce the computational complexity. In the case of more serious nonuniform sampling compared with the matched filter bank algorithm based on Doppler spectrum reconstruction, the proposed method can suppress azimuth ambiguities more effectively and shows potential in achieving wider range swath. In addition, the method can suppress the noise and sidelobe and improve the target to background ratio, which means the imaging performance is improved. The effectiveness of the proposed method is verified by simulation and actual data experiments.