摘要多输入多输出（Multiple-Input Multiple-Output，MIMO）技术可在频谱资源有限且信道环境复杂的情况下利用空间复用技术来提高系统的频率效率，利用空间分集技术提升误码性能对抗信道衰落。有限速率反馈技术就是接收端通过反馈链路向发送端传送有限的信道状态信息，从而可适度提高MIMO系统的误码性能和速率。

本文主要研究基于码本的有限速率反馈预编码技术。主要包括以下内容：第一，探讨了如何构建码本以及如何选出最优码字，本文主要介绍了DFT矩阵、Grassmannian子空间分组、Lloyd矢量量化三种码本构建方法以及基于最大似然（ML，Maximum Likelihood）接收机、基于线性接收机的码字选择算法。第二，比较了窄带不相关信道下，三种码本算法的性能，获得最优的码本算法为Lloyd矢量量化算法；第三，在平坦块衰落环境下，比较了不同反馈比特数对系统误码率的降低情况，发现反馈比特数越大，系统性能越好，但当反馈比特数超过6比特时，反馈比特数的增加对系统性能的提高作用可忽略。第四，比较了各种检测算法在无预编码和预编码MIMO系统中的性能，基于格基规约的检测算法相比于传统的线性检测算法对系统性能有更大的提升作用，其相比于ML检测算法既简单又便于实现。62966

毕业论文关键词  MIMO系统  有限反馈  预编码   格基规约   检测算法     

毕业设计说明书（论文）外文摘要

Title        Codebook-based    Limited   Feedback     Techniques    in    MIMO   Systems                                                  

Abstract    In the limited spectrum resources and the complex channel environment, MIMO technology can improve spectral efficiency by spatial multiplexing, and enhance the reliability of signal transmission and reduce channel fading through spatial persity techniques. Limited feedback technology is that the receive end sends limited channel state information to the sender through the feedback link, which can greatly enhance the performance of error and rate in MIMO systems.

    This paper studies the codebook-based limited feedback precoding techniques.The main work is as follows:  firstly, we discuss how to build codebook and how to select the optimal codeword, and this paper introduces the DFT matrix , Grassmannian subspace grouping , Lloyd vector quantization to build codebooks, and the algorithms based on ML receiver and linear receiver to select codeword; secondly, we compared the performance of the three kinds of codebook algorithms in the narrow-band irrelevant channel, and concluded that the optimal codebook algorithm is the Lloyd vector quantization algorithm in this case;  thirdly, we compared how different numbers of feedback bits to decrease system’s BER in flat block-fading environment, and find that the larger the number of feedback bits, the better the system’s performance, when the number of feedback bits exceeds 6 bits, the role to enhancement of system’s performance due to the increase of the number of feedback bits can be ignored; finally, we compared performance of various detection algorithm performance in the no pre-coding and pre-coding MIMO systems，and we can find that the detection algorithm based on lattice reduction improves system performance more compared to the traditional linear detection algorithms , and is simpler and easier to be achieved compared to the ML.

Keywords  MIMO system   Limited feedback   Precoding   Lattice Reduction   Detection algorithm

 目   次