本文为沙特阿卜杜拉国王科技大学（作者：Seifallah Jardak）的硕士论文，共75页。

由于多输入多输出（MIMO）雷达性能的提高，引起了研究者和工程师的广泛关注。由于MIMO雷达发射正交或部分相关的波形，这种新兴技术通过提供更好的参数可识别性、获得更高的空间分辨率、能够设计复杂的波束滤波器，因此优于相控阵雷达。为了避免干扰并提高信噪比，在给定感兴趣区域中最大化发射功率，并在其它区域最小化发射功率是非常有用的。该问题被称为发射波束设计，通常分为两步处理：首先通过最小化凸优化问题设计发射协方差矩阵，然后使用该凸优化表达式产生实际发射波形。

在本文中，我们提出了利用有限字母表常数和非常数包络符号产生相关波形的简单方法。为了产生有限的字母波形，提出的方法将高斯随机变量映射到相移键控、脉冲幅度和四边形幅度调制方案上。对于这种映射，高斯随机变量的概率密度函数被划分为M个区域，其中M是相应调制方案中的字母表个数。利用映射函数，推导出了高斯符号与有限字母符号之间的互相关关系。

论文的第二部分介绍了目标参数估计问题。为了确定目标的反射系数、空间位置和多普勒频移，采用最大似然估计的方法得到最佳性能。然而，最大似然估计涉及到一个两维搜索的问题，因此算法计算量巨大。因此我们提出了一种降低算法复杂度的次优算法，采用二维快速傅里叶变换联合估计空间位置和多普勒频移。为了评估本文所提出的估计器性能，推导了CRLB。仿真结果表明，所提出估计器的均方估计误差达到了CRLB的水平。

Thanks to its improved capabilities, theMultiple Input Multiple Output (MIMO) radar is attracting the attention ofresearchers and practitioners alike. Because it transmits orthogonal orpartially correlated waveforms, this emerging technology outperformed thephased array radar by providing better parametric identifiability, achievinghigher spatial resolution, and designing complex beampatterns. To avoid jammingand enhance the signal to noise ratio, it is often interesting to maximize thetransmitted power in a given region of interest and minimize it elsewhere. Thisproblem is known as the transmit beampattern design and is usually tackled as atwo-step process: a transmit covariance matrix is firstly designed by minimizinga convex optimization problem, which is then used to generate practicalwaveforms. In this work, we propose simple novel methods to generate correlatedwaveforms using finite alphabet constant and non-constant-envelope symbols. Togenerate finite alphabet waveforms, the proposed method maps easily generatedGaussian random variables onto the phase-shift-keying, pulse-amplitude, andquadratureamplitude modulation schemes. For such mapping, the probabilitydensity function of Gaussian random variables is divided into M regions, whereM is the number of alphabets in the corresponding modulation scheme. Byexploiting the mapping function, the relationship between the cross-correlationof Gaussian and finite alphabet symbols is derived. The second part of thisthesis covers the topic of target parameter estimation. To determine thereflection coefficient, spatial location, and Doppler shift of a target, maximumlikelihood estimation yields the best performance. However, it requires a twodimensional search problem. Therefore, its computational complexity isprohibitively high. So, we proposed a reduced complexity and optimumperformance algorithm which allows the two dimensional fast Fourier transformto jointly estimate the spatial location and Doppler shift. To assess theperformance of the proposed estimators, the Cram′er Rao Lower Bound (CRLB) isderived. Simulation results show that the mean square estimation error of theproposed estimators achieve the CRLB.

1 引言

2 发射波束形成与波形设计

3 目标参数估计

4 结论与评论

下载英文原文地址：

http://page5.dfpan.com/fs/4lc0j2a21929e16edf7/

更多精彩文章请关注微信号：