干扰背景下的信号探测

发布时间：2018-01-31 12:11

本文关键词： 阵列信号处理强干扰空间谱估计 DOA 正交投影变换扩展噪声子空间　出处：《东南大学》2017年硕士论文　论文类型：学位论文

【摘要】：在经典的信号处理领域中,阵列信号处理是一个非常重要的分支,具有极其重要的地位。近些年来,随着阵列信号处理的不断发展以及微弱目标源信号参数跟踪和估计方法研究的不断深入,空间谱估计,即波达方向(DOA,Direction of Arrival)估计已经成为一个重要的研究方向,在雷达、声呐、射电天文学和无线通信等国防和民用领域中具备重大应用价值。然而,在实际的阵列信号处理系统工作环境中,由于入射杂波信号的来源十分复杂,微弱的目标源信号往往湮没在强干扰信号和较强的背景噪声中,从而严重抑制了常规的空间谱估计方法在DOA估计过程中的性能。本文针对常规空间谱估计方法在强干扰、弱信号背景下DOA估计性能受到严重影响的问题展开了较为深入的研究,主要的工作内容和创新可以概括为如下几点:1.归纳和分析了常规背景下的空间谱估计方法,并且对常规空间谱估计方法中经典的Bartlett算法(常规波束形成法)、最小方差无失真响应(MVDR,Minimum variance distortion-less response)算法和多重信号分类(MUSIC,multiple signal classification)算法进行了不同条件下的仿真实验以及不同算法之间的性能比较。2.研究了两类适用于强干扰、弱信号背景下的空间谱估计方法:基于正交投影变换的DOA估计方法和基于扩展噪声子空间的DOA估计方法。(1)在基于正交投影变换的DOA估计方法中,假定预先已知强干扰信号来波到达方向的先验信息。在这样的背景下,通过求解干扰信号导向矢量张成子空间的正交投影,就可以直接对接收到的快拍数据样本进行预处理,将其投影到干扰信号的垂直子空间中,从而彻底消除来波信号强干扰的影响。于此同时,本文中定义了抑制因子,巧妙的避免了正交投影的过程对于微弱信号DOA探测的影响。在消除来波信号数据中强干扰信号的分量后,经典的空间谱估计方法的性能就不会受到干扰的显著抑制。(2)在基于扩展噪声子空间的DOA估计方法中,解决强干扰的技巧是通过将强干扰导向矢量所张成的信号子空间融入到原有的来波信号噪声子空间中,从而组成扩展的噪声子空间,以达到消除强干扰对弱目标源信号的影响。最后,通过常规的空间谱估计方法,例如MUSIC算法,就可以获取弱目标源信号DOA信息的估计值。本文分别对这两种方法进行了仿真实验和结果分析,比较了在不同干扰个数、不同阵元数、不同强弱信号相距角度等情况下这两种方法的性能。根据仿真实验的结果可以得出结论:无论是正交投影变换法还是扩展噪声子空间法,两者都能很好的解决在强干扰信号背景下微弱目标源信号DOA参数检测的问题。且知,正交投影变换法运算量适中但需要知道强干扰信号的先验信息,而扩展噪声子空间法无需预知强干扰的先验信息,适用范围更为广泛。3.分析了宽带信号背景下空间谱估计的基本原理。着重介绍了宽带信号条件下,非相干子空间处理方法(ISM,Incoherent Signal-subspace Method),并将其与基于正交投影变换和扩展噪声子空间的空间谱估计方法相结合,成功推广到强干扰条件下宽带信号DOA估计问题的求解中。本文给出了宽带信号空间谱估计算法的详细推导,并对非相干子空间处理的方法进行了不同条件下的仿真实验和性能分析。
[Abstract]:In the field of signal processing in classical array signal processing is a very important branch, has an extremely important position. In recent years, with the continuous development of array signal processing and target tracking and weak signal parameter estimation method of the continuous deepening of research, spatial spectrum estimation, direction of arrival (DOA Direction, of Arrival) estimation has become an important research direction in radar, sonar, and has great application value for radio astronomy and wireless communications and other military and civilian fields. However, in the work environment of array signal processing systems, due to the incident wave signal source of clutter is very complex, the target source signal is often lost in strong interference strong signal and background noise, which seriously inhibited the conventional spatial spectrum estimation method in DOA estimation process. Based on the conventional spatial spectrum estimation method In the background of strong interference, weak signal DOA estimation performance is affected by the problems are deeply researched, the main work and innovation can be summarized as follows: 1. summarizes and analyzes the general background of spatial spectrum estimation method, and the conventional Bartlett spatial spectrum estimation algorithm in classical methods (conventional beamforming method), minimum variance distortionless response (MVDR, Minimum variance distortion-less response) algorithm and multiple signal classification (MUSIC, multiple signal classification) algorithm are compared by simulation experiments under different conditions and the performance of different algorithms of.2. between the two kinds of suitable for strong interference, weak signal under the background of spatial spectrum estimation methods: orthogonal projection transform and DOA estimation method based on extended noise subspace estimation method based on DOA. (1) based on orthogonal projection transform DOA estimation In the method, the strong interference signal assumed known in advance to the direction of arrival of prior information. In this context, through the orthogonal projection method for solving the interference signal steering vector Zhang He space, you can directly to the snapshot data received sample pretreatment, projected to the vertical interfering signal subspace, thus to completely eliminate the effects of wave interference. At the same time, this paper defines the inhibitory factor, overcomingsome process orthogonal projection influence on weak signal detection. DOA signal data of strong disturbance signal in the elimination of the component after estimation significantly inhibited the performance of the method will not be interfered with the classic spectrum space. (2) in the extended noise subspace DOA estimation method based on the technique is adopted to solve the interference signal of strong interference by a steering vector space into the original signal The noise subspace, which consists of extended noise subspace, in order to eliminate the influence of strong interference on the weak target signal. Finally, the conventional spatial spectrum estimation method, such as MUSIC algorithm, estimation can obtain weak target source signal DOA information value. This paper analyzes the two methods of simulation experiment the results and analysis, comparing the different number of interference, different element number, different from the performance angle signal under the condition of the two methods. According to the simulation results it can be concluded that both the orthogonal projection method is also extended noise subspace method, both can well solve the problem of detection weak target source signal DOA parameters in strong interference background signal. And the orthogonal projection transformation method, the computational complexity is moderate but need to know the strong interference signal prior information, and the extended noise subspace method to predict strong Interference of prior information, a wider scope of application of.3. analysis of the basic principle of broadband signal under the background of spatial spectrum estimation. Introduced the broadband signal under the condition of non coherent subspace method (ISM, Incoherent, Signal-subspace and Method) and the orthogonal projection transform and extended noise subspace based spatial spectrum estimation method the combination successfully extended to broadband DOA signal under strong interference estimation problem. This paper gives a detailed estimation algorithm is derived and the wideband signal spatial spectrum, non coherent subspace processing in the simulation experiment and performance analysis under different conditions.

【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911.7

【相似文献】