宽带信号高分辨DOA估计算法研究

发布时间：2018-08-06 16:35

【摘要】：波达方向(DOA)估计是阵列信号处理中的一个重要分支,其在移动通信、电子侦察、无线电频谱监控等领域得到了普遍的应用。其中,对窄带信号DOA估计的基本算法的研究已日臻完善。但对于在雷达领域有着重要应用价值的宽带信号,其波达方向估计一直都是阵列信号处理领域的重点和难点。因此,宽带信号的DOA估计已成为近年来阵列信号处理中一个热门的研究内容。本文首先介绍和分析了现有的几种窄带子空间类DOA估计算法,并对相关算法的估计性能做了比较分析。在窄带子空间类DOA估计理论的基础上,针对越来越广泛使用的宽带信号,研究其高分辨DOA估计算法。文中介绍了两种典型的宽带信号DOA估计算法——ISM算法和CSM算法。对于CSM算法,本文介绍了其基本理论,并描述了聚焦矩阵和聚焦频率的选择问题。文中着重对ISM算法进行了理论分析和研究,并针对传统ISM算法中子频带的能量分布不均和信噪比各不相同而导致的算法估计性能不佳的问题,从算法运算量和低信噪比下算法性能的角度出发,对传统ISM算法进行改进,从而达到以较少的计算量或在较低信噪比条件下准确估计宽带信号源的方位。改进的内容如下:1、给出了一种新的宽带DOA估计算法——root-ISM算法,以多子带求根的思想实现DOA估计,从而有效地降低原ISM算法的计算量;2、在root-ISM算法的基础上给出root-WISM算法,该算法以子频带能量对DOA估计结果的加权平均取代子频带DOA估计结果的简单算术平均,改善root-ISM算法在信噪比较低时的估计性能;3、给出另一种在root-ISM算法基础上改进的f-root-WISM算法,以最大能量频点子频带估计取代多子频带综合估计,进一步降低算法的计算量。针对上述思路对相应算法进行了理论分析和仿真,从而验证了root-ISM算法及其两种改进算法的有效性和优越性。另外,本文介绍了小波去噪理论,并将其中的小波阈值去噪理论运用于宽带DOA估计系统中,对ISM算法在低信噪比下的估计性能问题作了进一步的探讨,即通过在进行DOA估计之前增加一个小波去噪预处理过程,间接改善算法的估计性能。仿真实验表明该方案能改善ISM算法的分辨能力并解决低信噪下易出现伪峰的问题。
[Abstract]:Direction of arrival (DOA) estimation is an important branch of array signal processing. It has been widely used in mobile communication, electronic reconnaissance, radio spectrum monitoring and other fields. Among them, the basic algorithm of narrowband signal DOA estimation has been improved day by day. However, for wideband signals with important application value in radar field, the DOA estimation is always the focus and difficulty in array signal processing. Therefore, DOA estimation of wideband signals has become a hot topic in array signal processing in recent years. In this paper, we first introduce and analyze several existing narrowband subspace-like DOA estimation algorithms, and compare and analyze the estimation performance of the related algorithms. Based on the theory of narrowband subspace-like DOA estimation, the high-resolution DOA estimation algorithm is studied for wideband signals which are more and more widely used. This paper introduces two typical DOA estimation algorithms for wideband signalsISM algorithm and CSM algorithm. In this paper, the basic theory of CSM algorithm is introduced, and the selection of focusing matrix and frequency is described. This paper focuses on the theoretical analysis and research of the ISM algorithm, and aims at the problem of poor estimation performance caused by the uneven distribution of energy in the neutron frequency band and the different signal-to-noise ratio (SNR) of the traditional ISM algorithm. From the point of view of computational complexity and performance of the algorithm under low SNR, the traditional ISM algorithm is improved to estimate the azimuth of the wideband signal source accurately with less computation or lower signal-to-noise ratio (SNR). The improved contents are as follows: 1. A new wideband DOA estimation algorithm, root-ISM algorithm, is presented. The DOA estimation is realized by using the idea of multiple subbands to find the root, which effectively reduces the computational complexity of the original ISM algorithm. On the basis of the root-ISM algorithm, the root-WISM algorithm is given. This algorithm replaces the simple arithmetic average of the subband DOA estimation result with the weighted average of the subband energy to the DOA estimation result, and improves the estimation performance of the root-ISM algorithm when the SNR is low. Another f-root-WISM algorithm based on the root-ISM algorithm is presented. In order to reduce the computational complexity of the algorithm, the multi-subband synthetic estimation is replaced by the maximum energy frequency point frequency band estimation. The theoretical analysis and simulation are carried out to verify the effectiveness and superiority of the root-ISM algorithm and its two improved algorithms. In addition, the wavelet denoising theory is introduced, and the wavelet threshold denoising theory is applied to the wideband DOA estimation system. The estimation performance of the ISM algorithm at low SNR is further discussed. In other words, the estimation performance of the algorithm is improved indirectly by adding a wavelet denoising preprocessing process before the DOA estimation. Simulation results show that the proposed scheme can improve the resolution of ISM algorithm and solve the problem that pseudo peaks are easy to appear under low signal noise.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7

【相似文献】