【作者】 贾寅修；

【导师】 常文革；

【作者基本信息】 国防科学技术大学 ， 信息与通信工程， 2006， 硕士

【摘要】 近年来,地表穿透合成孔径雷达(GPSAR)得到了广泛使用和飞速发展。作为一种近场探测SAR,其不仅容易受到各种射频干扰(RFI)的影响,还不可避免的要受到耦合杂波的干扰。国防科技大学超宽带雷达技术研究室在“十五”期间研制了轨道地表穿透合成孔径雷达(Rail-GPSAR),各种射频干扰和耦合干扰严重影响了Rail-GPSAR的探测性能。能否有效抑制这些干扰,是提高GPSAR成像质量和系统性能的关键。针对这个实际情况,本文就Rail-GPSAR的干扰抑制问题展开了深入的研究。在对Rail-GPSAR干扰的深入分析和对常用的干扰抑制方法研究的基础上,重点讨论了三类适用于Rail-GPSAR的干扰抑制方法:自适应滤波法、特征子空间滤波法和二维频域法。本文详细介绍了这三类算法的原理及特点,详细讨论了滤波过程中各种参数的选取问题,并对干扰抑制结果进行分析评估,对比各种方法的优点及局限性。自适应滤波是一种重要的统计信号处理手段,Rail-GPSAR杂波成份复杂,非常适合采用自适应滤波的办法。无论是利用自适应抵消滤波法去除耦合干扰,还是利用自适应预测误差法抑制RFI,都可以取得良好的效果。特征子空间滤波法则是通过特征向量分解的方式,分别建立信号子空间与干扰子空间。通过将回波数据向信号子空间投影的方式滤除杂波。其特点是能够极大的抑制掉强干扰的频率成份。二维频域法的思想是利用有效目标回波与干扰杂波在二维频域内的支撑域不同的特点,比较准确的进行谱估计。本文采取一套行之有效的信号处理流程,有效的抑制了干扰杂波。通过对Rail-GPSAR实测数据的处理结果进行评估分析,发现上述三类算法非常适合Rail-GPSAR的干扰抑制,均能产生令人满意的效果。更多还原

【Abstract】 Recent years, Ground Penetrating Synthetic Aperture Radar(GPSAR) has been well developed and been applied to many fields. As a short-range detection radar, it is influenced not only by the Radio Frequency Interference (RFI), but also by the coupling clutter. To improve the SAR images quality, we have to suppress the clutter effectively.The paper is based on the subject of clutter suppression of the Rail-GPSAR system which was build by National University of Defense Technology. Here it introduces many of traditional approaches to suppress the clutter. The paper pays much attention to these approaches: Adaptive Filtering, Eigen-subspace Decomposition Filtering and 2D Fourier Transform Filtering.Adaptive filtering can not only cancel the coupling clutter but also predict the RFI. The approach proved to be effective to restrain the complex clutter of Rail-GPSAR.Eigen-subspace decomposition filtering suppresses the clutter by projecting the received signal to the‘signal-subspace’.‘Signal-subspace’is obtained by eigenvector-subspace decomposition.2D Fourier transform filtering is a novel way to suppress the interference. It projects the received signals to the 2D Fourier Domain. The target echoes and the interference clutter are distributed to different areas. Then we can estimate the spectrum of RFI and restrain the clutter by a wiener filter.By analyzing the result, we found these three approaches can suppress the clutter effectively. The performance of each filter is satisfied.更多还原