【作者】 焦计平；

【导师】 周又玲；

【作者基本信息】 海南大学 ， 通信与信息系统， 2011， 硕士

【摘要】 近年来,人们对通信需求不断提高,而无线通信也已经成为一种重要的通信方式.在无线通信中,智能天线的作用变得日益重要.智能天线在提高频谱利用率,增大系统容量和扩大信号覆盖范围等方面显示出巨大潜力,已经成为第三代移动通信系统特别是TD-SCDMA标准中的一项关键技术.基于智能天线的重要性,本文对智能天线进行了研究.首先介绍了智能天线的基本概念,包括天线电参数、方向图、阵列的空间结构和信号模型等.它们是研究智能天线的基础.接下来,对智能天线中的一些关键算法进行了研究,这也是本文的重点.所研究的关键算法主要为：波达估计算法和波束形成算法.主要的研究成果为：1研究了超分辨波达估计算法,对MUSIC算法进行了仿真.由于现有ESPRIT算法需要计算矩阵束的广义特征值,这使得算法的实现变得非常困难.本文针对这一缺陷提出了改进ESPRIT算法和ESPRIT谱,新的算法尽量避免了计算矩阵束的广义特征值；2根据线性方程组理论和最小二乘思想对基于Chebyshev多项式的方向图综合算法的计算方式进行了改进,改进的算法避免了参数映射时复杂的三角函数运算；3研究了均匀直线阵与FIR滤波器的相似性,提出了基于FIR滤波器的波束形成算法；4将矩阵求逆引理应用于MV准则中,提出了基于MV准则的自适应波束形成算法,新的算法避免了矩阵求逆运算；5将RBF神经网络应用于波束形成中,提出了基于RBF神经网络的波束形成算法,并且根据最小二乘思想计算出了RBF神经网络的等效权值矢量.另外,本文对改进算法进行了计算机仿真,仿真结果验证了改进算法的正确性及有效性.更多还原

【Abstract】 With growing demands for communication services, the wireless communication has become an important communication mode. Meanwhile, the smart antenna has been playing a increasingly significant role in wireless communication. It is strongly probable that the smart antenna will be used for improving the spectrum utilization ratio, increasing the systematic capacity, extending the signal coverage and so on. The smart antenna has become one of key technologies of 3G mobile communication systems, especially TD-SCDMA.Owning to their important roles, smart antennas are researched in this paper. First of all, the thesis introduces some basic concepts of the smart antenna, such as the basic antenna parameters, the pattern, the spatial structure of arrays, the signal model, and etc, which are the fundamentals of are antennas. Then some key algorithms of smart antennas have been researched, which is the main content of the thesis. These researched algorithms consist of DOA estimation algorithm and beam forming algorithms. Some results have been obtained in and are listed as follow.1. Some super-solution DOA estimation algorithms are researched. The MUSIC algorithm is simulated and the improved ESPRIT algorithm has been proposed. The improved algorithm overcomes some defects of ESPRIT algorithm and avoids finding out the generalized Eigen values of matrix pencil.2. The pattern synthesis algorithm based on the Chebyshev polynomial is improved by using the linear equations theory and the least squared method. The new algorithm avoids the complicated trigonometric transforms.3. The similarity of the uniform linear array and the FIR filter is researched and a beam forming algorithm based on FIR filter is proposed.4. The beam forming algorithm based on MV criterion has been proposed in accordance with the Matrix Inversion Lemma. The new algorithm avoids the matrix inversion.5. The beam forming based on RBF Artificial Neural Network is proposed and the equivalent weight vector is deduced according to the least squared method.Moreover, the improved algorithms are simulated and the simulation results show the correctness and effictiveness of these algorithms.更多还原