Massive MIMO导频设计与信道估计

发布时间：2019-01-02 07:21

【摘要】：多天线系统(MIMO)利用传输分集、空间复用等技术充分挖掘维度资源,提高传输效率和通信质量。随着通信技术的发展,4G蜂窝网络中的多用户MIMO并不能对频谱效率和能量效率有数量级的提升,并且为满足通信的大容量、低功耗、低成本要求,未来5G网络提出在基站端布置大量的天线,在相同的时频资源块上服务多个小区用户,增加有用信号的功率,从而增加信干比,能够显著克服信道衰落和噪声的影响,使得基站处理能力得到显著提升。论文首先介绍了 TDD和FDD两种Massive MIMO帧结构,并阐述了 Massive MIMO使用TDD模式原因,然后对Massive MIMO TDD系统模型和导频污染进行介绍,详细分析了系统上行导频传输和信道估计,上行数据传输和MRC检测,下行数据接收过程,并搭建Massive MIMO系统仿真平台,给出仿真流程图分析了传统导频设计的LS估计和MMSE估计性能。然后在Massive MIMO系统框架下介绍了单小区半正交导频设计原理,提出半正交导频修正方案。给出修正后导频的设计原理和帧结构,并对导频设计的上下行传输过程以及性能进行研究,将其与传统导频和半正交导频进行性能对比,给出了导频设计方案仿真流程图,在之前建立的Massive MIMO仿真平台对不同导频设计方案进行性能对比。最后介绍两种多小区协作信道估计方式。第一种为基于Bayes估计的协作式信道估计。首先介绍Bayes估计原理,分析Bayes估计均方误差,到达角和协方差矩阵的影响,然后根据Bayes估计提出一种协作式信道估计策略,将用户分组,找到信道估计均方误差最小的一组用户同时进行信道估计;第二种为基于TCGTR的协作式信道估计,此方法为第二章单小区半正交导频设计的扩展,将其运用于多小区系统,对TCGTR估计过程进行了详细阐述;最后将两种估计方法进行仿真验证,其性皆优于传统信道估计方式。
[Abstract]:Multi-antenna system (MIMO) exploits dimensionality resources by means of transmission diversity and spatial multiplexing to improve transmission efficiency and communication quality. With the development of communication technology, multiuser MIMO in 4G cellular network can not improve spectrum efficiency and energy efficiency by an order of magnitude, and to meet the requirements of large capacity, low power consumption and low cost, In the future 5G network proposes to deploy a large number of antennas at the base station to serve multiple cell users on the same time-frequency resource block to increase the power of useful signals, thus increasing the signal-to-interference ratio, which can significantly overcome the influence of channel fading and noise. So that the base station processing capacity has been significantly improved. In this paper, two kinds of Massive MIMO frame structures, TDD and FDD, are introduced, and the reason why Massive MIMO uses TDD mode is expounded. Then, the Massive MIMO TDD system model and pilot pollution are introduced, and the uplink pilot transmission and channel estimation are analyzed in detail. Uplink data transmission, MRC detection, downlink data receiving process, and Massive MIMO system simulation platform are built. The simulation flowchart is given to analyze the LS estimation and MMSE estimation performance of the traditional pilot design. Then, the design principle of semi-orthogonal pilot in single cell is introduced under the framework of Massive MIMO system, and a semi-orthogonal pilot correction scheme is proposed. The design principle and frame structure of modified pilot are given, and the transmission process and performance of pilot design are studied. The performance of pilot design is compared with that of traditional pilot and semi-orthogonal pilot, and the simulation flow chart of pilot design scheme is given. The performance of different pilot design schemes is compared with the previous Massive MIMO simulation platform. Finally, two methods of multi-cell cooperative channel estimation are introduced. The first is cooperative channel estimation based on Bayes estimation. Firstly, the principle of Bayes estimation is introduced, and the effects of mean square error, angle of arrival and covariance matrix of Bayes estimation are analyzed. Then, a cooperative channel estimation strategy based on Bayes estimation is proposed to group users. A group of users with the least mean square error of channel estimation is found to estimate the channel at the same time. The second is cooperative channel estimation based on TCGTR. This method is an extension of single cell semi-orthogonal pilot design in chapter 2. It is applied to multi-cell system and the process of TCGTR estimation is described in detail. Finally, the two estimation methods are verified by simulation, and their performance is better than the traditional channel estimation method.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN919.3

【参考文献】