【作者】 孟祥东；

【导师】 保铮；

【作者基本信息】 西安电子科技大学 ， 信号与信息处理， 2009， 博士

【摘要】 空时二维自适应信号处理技术(Space-Time Adaptive Processing)能有效地改善机载相控阵雷达对地杂波抑制的性能。这项技术已经逐渐应用到机载雷达对空中和地面运动目标检测中,尤其是在机载预警雷达上的应用。在复杂环境下要检测运动目标我们必须对地杂波进行有效地抑制。由于载机平台的运动,机载相控阵雷达的地杂波谱在多普勒上表现出主杂波展宽和旁瓣杂波扩散,并且是空间和时间二维耦合的。传统的一维时域或空域滤波器均不能形成与地杂波相匹配的凹口,达到有效抑制杂波的目的。空时二维自适应信号处理是在空间和时间上的二维联合滤波,其滤波权系数是根据最大信杂噪比准则计算得到的。由于同时具有空间和时间自由度,它可以在二维谱平面上形成凹口有效地抑制地杂波,同时保证目标信号获得足够的增益。因此空时二维自适应滤波就成为了机载相控阵雷达抑制杂波的一项关键技术。考虑到实际环境和系统的复杂性,空时二维自适应处理要实际应用还有一系列的问题需要解决。本论文针对空时二维自适应处理技术在机载相控阵雷达中应用所遇到的工程实际问题以及机载雷达对运动目标检测的相关问题展开研究,提出和改进了一些信号处理算法,并针对一些具体问题给出了解决方案。本论文的研究内容主要包括以下几个方面:1.第二章系统地研究和分析了机载预警雷达非正侧面阵杂波的分布与系统参数的关系,分析了杂波距离非平稳性产生的主要原因。近程杂波的距离非平稳性是影响杂波抑制和动目标检测的主要因素。在准确掌握机载非正侧面阵雷达杂波谱分布规律的情况下提出了两种消除近程杂波影响的方案,用以解决杂波距离非平稳性的问题。一种方法是利用阵列天线俯仰向的阵元(或子阵)做俯仰向空域滤波抑制近程杂波以削弱杂波的距离非平稳性,其滤波权系数是由载机平台位置参数和雷达系统参数开环计算得到的。另一种方法是根据已知的雷达系统参数计算出近程杂波的距离多普勒二维分布,在STAP处理时将包含近程杂波的距离单元剔除使之不参与协方差矩阵的统计,这样就减少了样本污染对STAP的影响。这两种方法都是非自适应的,在系统误差较小的情况下可以达到很好的效果。2.第三章考虑实际系统中误差较大的情况,对俯仰向空域滤波的方法进行改进,提出了自适应俯仰滤波抑制地杂波的方法。对俯仰向自适应滤波权系数的计算给出了两种选样本的方式,一是在多普勒滤波后的杂波支撑区内,取多普勒通道的数据作训练样本,二是直接在脉冲域选样本。本章将先俯仰向自适应滤波后STAP处理的方法与现有的三维STAP处理的方法做了对比,实验结果证明先俯仰向滤波后STAP的方法性能较好而且运算量小。3.第四章研究了重叠子阵合成的降维STAP方法,对滑窗重叠子阵合成的降维方式进行了改进,提出了阵列子阵的合成方式和保留辅助列子阵的合成方式。利用了保留辅助列子阵的合成方式提出了干扰与杂波级联抑制的空时二维自适应处理方法。即先利用大子阵和辅助列子阵做空域自适应处理抑制干扰,然后再做空时二维自适应处理抑制杂波。这样既降低了空时自适应处理的维数又同时抑制了干扰。通过仿真实验证明了这种算法的性能与相同空域自由度情况下干扰和杂波同时抑制的STAP性能接近,而运算量却大大减少。4.第五章研究了在机载双基雷达上应用STAP技术抑制杂波的方法。对机载双基雷达杂波二维分布与双基几何构型的关系进行了分析,推导出了一个计算机载双基雷达杂波二维分布的公式,归纳出了机载双基雷达杂波谱的一些变化规律。通过研究发现,要利用STAP技术有效抑制机载双基雷达的地杂波,必须解决双基雷达杂波的距离非平稳性问题,因此提出了一种先利用阵列天线的空域自由度解距离模糊,后进行角度多普勒频移补偿的方法。这种方法有效地解决了在距离模糊的情况下消除杂波距离非平稳性的问题,为STAP技术在机载双基雷达上的应用打下了基础。5.第六章对影响SAR动目标检测的图像对的相干性做了研究。SAR图像对的相干性直接影响了雷达对地面动目标检测的性能。影响SAR图像相干性的因素有很多,本论文着重研究了基线和地形坡度对SAR图像相干系数的影响,利用SAR成像的基本原理,从数值上分析了基线、地形坡度等因素对相干系数的影响。针对两种不同的发射信号,分析推导出了计算相干系数的公式。本论文所给出的计算相干系数的公式对SAR运动目标检测的基线设计有参考意义。更多还原

【Abstract】 Space Time Adaptive Processing(STAP)can effectively improve the performance of ground clutter suppression in airborne phased array radar. This technique has been gradually applied to airborne radar for air moving targets indication (AMTI) and ground moving targets indication (GMTI), especially for airborne early warning (AEW) radar.The ground clutter has to be suppressed effectively before moving target indication. Due to the motion of airborne platform, the ground clutter spectra for airborne phased array radar are distributed in space-time domain. But the traditional one-dimensional filter can not form a notch that matches for the ground clutter, so it can not suppress clutter effectively. However, Space Time Adaptive Processing is a two-dimensional filter in space-time domain, whose weight is calculated by maximum SCNR criterion,so it can suppress ground clutter perfectly. Therefore, STAP becomes a key technique in airborne phased array radar for clutter suppression.Focusing on the engineering techniques for STAP applying to airborne phased array radar and the related problems of airborne radar for MTI, the dissertation presents and improves some signal processing algorithms. The main contents of the dissertation are described as following.1. Chapter 2 specifically analyzes the relationship between the clutter distribution of AEW radar and its radar system parameters, and finds out the causation of clutter range dependence. The range dependence of short-range clutter is the key factor affecting clutter suppression and MTI. Since the clutter distribution law for airborne non-side-looking array radar has been fully researched, we propose two approaches to mitigating the effect of short-range clutter. The first one is called elevation filtering approach, i.e. the short-range clutter is suppressed by utilizing the elevation elements of array antenna. The weight for elevation filtering is calculated nonadaptively. The second one is named elimination approach, i.e. the clutter range-Doppler distribution is firstly calculated in accordance with radar system parameters, and the range cells, including the short-range clutter, are eliminated before estimating covariance matrix. By doing so, the effect of training data pollution in STAP is mitigated. Both approaches are realized nonadaptively, and they can achieve good performance when the system error is negligible.2. Chapter 3 improves the elevation filtering approach, and presents an approach to the calculation of the elevation weight adaptively. This dissertation proposes two ways of training data strategies: in the first way, the training data is selected from the Doppler domain in the clutter support area, while in the second one, the training data is directly selected in the pulse domain. The performance of elevation filtering approach and 3-D STAP are compared, and the simulation experiment testifies that the elevation filtering STAP is superior to 3-D STAP in both performance and computational load.3. The reduced-dimension STAP related to overlapped-subarray synthesis is studied in Chapter 4. The dissertation contributes to this study by improving the sliding overlapped-subarray synthesis. The planar subarray synthesis and the subarray synthesis with reserved auxiliary column-subarray are presented in this dissertation. For the subarray synthesis with reserved auxiliary column-subarray, a cascaded approach in which the interference and clutter are suppressed respectively is proposed. First, the interference is suppressed adaptively by utilizing one main subarray and all the auxiliary column-subarrays, and then the clutter is suppressed by STAP. The dimension for STAP can be reduced at the same time as interference suppression, and the simulation experiment testifies that with the same spatial DOF, the performance of this cascaded algorithm approaches the performance of STAP in which interference and clutter are suppressed simultaneously. However, the computational load is reduced greatly.4. Chapter 5 studies the STAP application to bistatic airborne radar. This dissertation specifically analyzes the relationship between the bistatic clutter distribution and the bistatic geometry, and then deduces a formula for calculating the bistatic clutter distribution. Moreover, we find out several change laws of bistatic clutter spectra. In the author’s opinion, the range dependence of bistatic clutter has to be mitigated before STAP technique can be applied to bistatic airborne radar, therefore, we suggest that the clutter range ambiguity should be resolved by utilizing spatial DOF (degrees of freedom) firstly, and then the clutter range dependence is compensated in angle-Doppler domain. This approach effectively solves the problem of bistatic clutter range dependence in the case of range ambiguity, so it is of great help for STAP applying to bistatic airborne radar.5. Chapter 6 mainly studies the coherence of SAR image pairs. The coherence of SAR image pairs affects the performance of GMTI directly. There are many factors affecting the coherence of SAR image pairs, but this dissertation mainly focuses on the effect of baseline and ground slope. In accordance with SAR imaging principle, we study the relationship between coherence coefficient and baseline numerically. The presented formula for coherence coefficient calculation is of significance to baseline designing for SAR-GMTI.更多还原