【作者】 李化；

【导师】 王华奎；

【作者基本信息】 太原理工大学 ， 电路与系统， 2011， 博士

【摘要】 MIMO技术可提高频谱利用率,在不增加带宽的情况下成倍地提高通信系统的容量,这在频谱资源日益紧张的今天具有非常重要的意义。MIMO技术被认为是实现高速宽带无线互联网的关键技术之一,在新一代移动通信系统中有广阔的应用前景。MIMO通信系统的性能很大程度上取决于无线信道的衰落特性。故需要对无线MIMO衰落信道的模型进行研究,建立参数化的便于描述的无线衰落信道模型。在MIMO系统中,信道数量的增多、发送信号空间维的扩展使信道估计己成为制约系统性能和实际应用的关键技术。本文主要对无线MIMO系统中的信道模型、信道估计进行了研究。本文的主要内容如下：介绍了多输入多输出无线通信的研究背景和基本概念,阐述了在MIMO无线通信领域中正受到关注的和仍有待解决的基本问题,指出了本文的研究方向。分析了电波传播的基本机制,研究了无线信道的特征。为无线信道建模奠定了理论基础。研究了无线MIMO系统的系统模型和信道容量。针对MIMO无线信道建模这一具有挑战性的前沿课题,在研究平坦衰落信道模型、频率选择性衰落信道模型和空间信道典型模型的基础上,提出了一个具有直观物理意义的时变相关MIMO信道模型。该模型考虑了传播环境对信号产生的各种影响,并具有较低的计算复杂度；并建立了一个MIMO信道快速仿真器。以平坦衰落窄带信道为背景,基于无线MIMO传输帧结构和数学模型,利用训练序列推导了MIMO信道系数的最大似然估计值,最小二乘估计值和最小均方误差估计值。基于信道阶数为L,FIR结构的频率选择性信道,提出了新的基于叠加训练序列的信道估计方法,通过在信息序列上叠加训练序列,仅利用一阶统计量的方法实现信道估计。此算法针对时变信道中可利用的观测数据有限的情况,利用时间上相邻的信道系数之间的相关性,在没有带宽损失的情况下可以估计出信道系数。并在理论上证明了此算法的无偏性和有效性。以有理空间理论为基础,在信道阶数己知条件下,研究了基于子空间方法的MIMO盲信道估计算法。将此算法与HGA相结合,提出了基于HGA的MIMO盲信道估计算法。将待估计的信道系数和阶数阶数采用不同的编码方式编码在同一染色体中,以基于子空间的MIMO盲信道估计算法中得到的关于信道系数的方程式为基础构造适应度函数,同时估计出信道系数和信道阶数。解决了以往算法中需将信道阶数定义为已知或设定信道阶数上界的难题。仿真结果显示本文提出的算法有较快的收敛速度,其估计误差和基于子空间方法的MIMO盲信道估计算法相当。更多还原

【Abstract】 By adopting MIMO (Multiple-Input Multiple-Output) technology, the capacity and spectrum efficiency of wireless communication systems can be increased significantly without the expense of bandwidth. It is believed that the MIMO technology will be one of the key technologies that will be uesd in the high-speed broadband wireless Internet access networks and has wide application prospect in new generation mobile communications in the future.The performances of the MIMO systems are decided by the fading characteristics of the wireless channel to a great extent. So, it needs to research the modeling and simulation of MIMO channels. The parametrical models of MIMO fading channels need to be established. In MIMO systems, with the increasing number of channel parameters and the extension of transmit signals in space dimension, the channel estimation have become the bottleneck of system performance and practical application. In this paper, I have mainly made further research on the related theories in the channel models and channel estimation and implementation algorithms on the base of others’research work. The main contents of the paper are expressed as follows:The background and basic concept of MIMO techniques are present firstly. A variety of concentrated problems and fundamental questions are introduced. The research directions are indicated later.The basic mechanism of radio wave propagation and characteristic of wireless channel are analyzed. The system models and capacity of MIMO are study based on the system models and capacity of SISO. The numerical analysis to MIMO channels capacity was carried on by simulation.MIMO wireless channel modeling is studied systematically. Above all, the flat fading channel models, the frequency-selective channel models and typical space-selective channel models were studied. Second, putting forward the methods of generating time varied correlation MIMO channels model. This model has considered a variety of contribution from propagation environment. Third, putting forward a fast MIMO channel simulator This improved the accuracy and saved the time of simulation.The MIMO channel estimation base on the training sequence is studied under flat narrow band channel. The channel parameters are deduced by different rule, just as maximum likelihood estimate (MLE), least square estimate (LSE) and minimum mean square error estimate. The new MIMO channel estimation based on the superimposing training sequence was put forward. The channel parameters are deduced by superimposing training sequence on the signal sequence, utilizing first order statistic. Aim at limited observation data in time varied channel, this new algorithm is bring forward. Unbiasedness and effectiveness have been verified.MIMO blind channel estimation base on subspace is studied base on rational space. A MIMO blind channel estimation base on HGA is presented. In this algorithm, channel parameters and order are coding in same one chromosome. The fitness function is constructed by MIMO blind channel estimation base on subspace. Channel parameters and order are estimated at same time. It solves the puzzle that the order needs to be assigned. The result of simulation displays:this new algorithm have faster convergence rate and its estimation error is correspond to MIMO blind channel estimation base on subspace.更多还原