导向矢量失配条件下的稳健自适应波束成形研究

发布时间：2018-05-23 14:00

本文选题：数字天线阵列 + 导向矢量　；参考：《中国科学技术大学》2014年博士论文

【摘要】：近年来数字天线阵列技术在现代国防军事工业、移动通信和声纳等领域发挥了越来越重要的作用。自适应波束成形作为数字阵列的核心技术,它的不断发展极大的推动了数字阵列的研究。与数据独立型波束成形方法相比,自适应波束形成具有更高的分辨率和更强的干扰抑制能力,并且这些特性都是建立在期望信号导向矢量等信息精确已知的前提下。但是,与传统的相控阵一样,数字阵列在工作时,也面临了诸多误差因素的影响,例如,阵元间互耦、幅相误差、阵元位置误差等。而一般的自适应波束成形方法对误差因素造成的导向矢量失配是比较敏感的。因此,近年来,大量的研究工作将重心放在了如何提高自适应波束形成器在导向矢量失配条件下的稳健性上。本文在已有工作的基础上,研究了几种新的稳健自适应波束成形方法。本文首先在导向矢量的椭球不确定集基础上,以最小化误差敏感性为目标来设计稳健自适应波束形成器。与传统的输出性能最佳化法相比,误差敏感性最小化法在导向矢量失配程度一定的前提下,对算法中导向矢量误差范数上界的取值不敏感。理论上,误差敏感性最小化法仍然属于基于导向矢量不确定集的稳健自适应波束成形。但是,对于这一大类稳健波束成形方法来说,当参考导向矢量与实际的导向矢量之间的失配程度较高时,其输出信干噪比会出现较为明显的下降。接着,针对以上所述方法的输出性能在较大导向矢量失配度下出现下降的问题。本文利用特征子空间分解定理构造了关于导向失配度的估计方程,并以这些方程为基础发展出了具有自适应可调误差半径的迭代稳健自适应波束成形技术。在每一步迭代中都可以依据一定的准则估计出导向矢量不确定集的半径。随后以该误差半径为基础,求解出相应的最优导向矢量,并将其作为下一步迭代中使用的参考向量。该处理方法能够有效提高较大导向矢量失配度下自适应波束形成器的输出性能。另一方面,该迭代搜索法也可被推广到椭球不确定集中。但是,以上的迭代处理法在改善稳健性的同时也增加了计算的复杂度。所以,本文又提出了一类新的基于泰勒级数展开的迭代稳健自适应波束成形。该方法无需构建导向矢量的不确定集,其在输出性能和处理复杂度之间进行了一定的平衡。最后,考虑到均匀矩形平面阵中阵元间互耦的对称性,本文进一步研究了具有低互耦敏感度的稳健自适应波束成形。通过在均匀矩形面阵的四周添加若干层辅助阵元,可以在不构建导向矢量不确定集的前提下,也能够有效改善自适应波束形成器的性能。但是辅助阵元的引入也会造成在特定互耦系数条件下的空间谱估计中出现盲角问题,该问题可以通过使用不维度的噪声子空间进行循环空间谱估计来解决。
[Abstract]:In recent years, digital antenna array technology has played a more and more important role in modern defense and military industry, mobile communication and sonar. As the core technology of digital array, adaptive beamforming has greatly promoted the research of digital array. Compared with the data independent beamforming method, adaptive beamforming has higher resolution and stronger interference suppression capability, and these characteristics are based on the premise that the information such as the desired signal guidance vector is accurately known. However, like the traditional phased array, the digital array also faces the influence of many error factors, such as mutual coupling between array elements, amplitude and phase error, position error and so on. The conventional adaptive beamforming method is sensitive to the mismatch of steering vector caused by error factors. Therefore, in recent years, a lot of research has focused on how to improve the robustness of adaptive beamformer under the guidance vector mismatch condition. Based on the previous work, several new robust adaptive beamforming methods are studied in this paper. In this paper a robust adaptive beamformer is designed based on the ellipsoidal uncertainty set of the guidance vector with the aim of minimizing error sensitivity. Compared with the traditional output performance optimization method, the error sensitivity minimization method is insensitive to the upper bound of the guidance vector error norm on the premise of a certain mismatch degree of the guide vector. In theory, error sensitivity minimization still belongs to robust adaptive beamforming based on guidance vector uncertainty set. However, for this class of robust beamforming methods, when the mismatch between the reference steering vector and the actual guidance vector is high, the output signal-to-noise ratio will decrease obviously. Then, the output performance of the method is reduced under the larger mismatch of steering vector. In this paper, the estimation equations of steering mismatch are constructed by using characteristic subspace decomposition theorem. Based on these equations, an iterative robust adaptive beamforming technique with adaptive adjustable error radius is developed. The radius of the guidance vector uncertainty set can be estimated according to certain criteria in each step iteration. Based on the error radius, the corresponding optimal guidance vector is solved and used as the reference vector in the next iteration. This method can effectively improve the output performance of adaptive beamformer with large steering vector mismatch. On the other hand, the iterative search method can also be extended to ellipsoidal uncertain sets. However, the above iterative methods not only improve robustness, but also increase computational complexity. Therefore, a new class of iterative robust adaptive beamforming based on Taylor series expansion is proposed. The method does not need to construct an uncertain set of guidance vectors, and it balances the output performance with the processing complexity. Finally, considering the symmetry of mutual coupling between elements in a uniform rectangular planar array, robust adaptive beamforming with low mutual coupling sensitivity is further studied in this paper. By adding several layer auxiliary elements around the uniform rectangular array, the performance of the adaptive beamformer can be improved effectively without constructing the guidance vector uncertainty set. However, the introduction of auxiliary elements will also lead to the blind angle problem in the spatial spectral estimation under the condition of specific mutual coupling coefficient. This problem can be solved by using non-dimensional noise subspace to estimate the cyclic spatial spectrum.
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN820.15

【共引文献】