1 / 62
文档名称:

MIMOSC-FDMA系统中半盲信道估计新方法研究.pdf

格式:pdf   页数:62
下载后只包含 1 个 PDF 格式的文档,没有任何的图纸或源代码,查看文件列表

如果您已付费下载过本站文档,您可以点这里二次下载

分享

预览

MIMOSC-FDMA系统中半盲信道估计新方法研究.pdf

上传人:cherry 2014/6/6 文件大小:0 KB

下载得到文件列表

MIMOSC-FDMA系统中半盲信道估计新方法研究.pdf

文档介绍

文档介绍:南京邮电大学
硕士学位论文
MIMOSC-FDMA系统中半盲信道估计新方法研究
姓名:刘元元
申请学位级别:硕士
专业:通信与信息系统
指导教师:宋荣方
2011-03
南京邮电大学硕士研究生学位论文摘要
摘要
多径性与时变性是无线信道的两大特点,特别是当系统采用空时编码时,接收端需要
准确知道信道特性才能进行有效的解码和相干解调,因此信道估计的准确性对无线系统是
极为重要的。信道估计方法可分为三类:盲信道估计,非盲信道估计和半盲信道估计。半
盲估计作为盲估计与非盲估计的折中方案,它利用尽可能少的导频信息,同时充分利用所
传输的用户信号的统计信息来优化跟踪信道参数。半盲估计方法性能可靠且占用很少的信
道资源,预计将为下一代移动通信系统所广泛采用。
由于有着较低的峰值平均功率比(PAPR),单载波频分多址(SC-FDMA)系统已经被
3GPP 确定为长期演进(LTE)上行链路的多址接入技术,它是正交频分多址(OFDMA)
的一种改进形式。本文提出一种应用于 MIMO SC-FDMA 系统的半盲信道估计方法,它联
合了基于训练的最小二乘(LS)准则以及基于线性预测的盲约束。仿真结果表明,在采用
相同数量的训练符号时,本文提出的方法明显优于传统 LS 算法。
I
南京邮电大学硕士研究生学位论文 Abstract
Abstract
The two major characteristics of the wireless channel are multi-path and time-variant
properties. For the performance of decoding and coherent demodulation, the receiver has to know
the accurate knowledge of the channel. Therefore, the accuracy of channel estimation is of crucial
importance to wireless system. Channel estimation methods can be divided into three categories:
the blind method, the non-blind method and the semi-blind one as bination of the first two
methods. With a small number of training symbols, the semi-blind method optimizes the
parameters of the tracking channel by using the statistics of the transmitted signals. The
semi-blind channel estimation has been recently considered as petitive method for the next
generation munications due to its reliability as well as using small channel resources.
Due to its low peak-to-average power ratio, single carrier frequency division multiple access
(SC-FDMA), a modified form of Orthogonal FDMA (OFDMA), has been adopted as the uplink
multiple access scheme in the Long Term Evolution (LTE) of cellular systems by 3GPP. In this
paper, a semi-blind channel estimation method is presented for multiple input multiple output
(MIMO) SC-FDMA systems. The new method uses a training-based least squares (LS) criterion
along with a blind constraint based on the linear prediction. Simulation results