使用週期平穩特性於多路徑通道之盲蔽式鑑定與等化

Blind Identification and Equalization of Multipath Channels using Cyclostationarity

邱志華(Jyu-Hwa Chiou)

指導教授 ： 黃正光

元智大學/工程學院/電機與資訊工程研究所/碩士(1995年)

摘要

數位通訊系統受到多路徑通道的影響,對信號接收造成碼間干擾(
Intersymbol Interference)和頻率選擇性衰減(frequency-selective
fading)的問題,長久以來即被工程師所研究探討.傳統用於估計衰減因子(
attenuation factor)和時間延遲(time delay)的方法常會受到雜訊和干
擾的影響,而發生較大的誤差.因此,本文利用大多數數位調變信號本身具
有週期平穩(cyclostationary)的特性,來做多路徑通道的參數估計,只要
原始訊號存在週期平穩的特性(具有週期頻率),即可利用週期自生相關函
數(cyclic autocorrelation function)的性質和頻譜相關密度函數(
spectral correlation density function)的性質,壓抑雜訊甚至其它干
擾信號,以提高性能.在本論文中提出盲蔽式通道鑑定(blind channel
identification)的方法,是不須藉由傳送training sequence而得知經過
通道之前的信號,只須知道發射訊號的調變方式,baseband波形與symbol
rate,吾人就可利用其統計特性相同的性質,求得其週期自生相關函數,然
後針對多路徑通道的問題,推導出通道參數估計的方法,以做到多路徑通道
盲蔽式鑑定與等化的目的.

關鍵字

週期平穩 ； 多路徑通道 ； 盲蔽式 ； 鑑定 ； 等化

並列摘要

Digital communication through channels with multipath
propagation is subject to intersymbol interference and
frequency-selective fading. The traditional methods used to
estimates the attenuation factors and time delays of the
multipath channel are usually very sensitive to the presence of
noise and interference.This thesis exploits the
cyclostationarity of most digital modulation signals for
estimating the parameters of multipath channels. The
identification algorithm is derived by using the second-order
cyclic autocorrelation functionand spectral correlation density
function. The proposed algorithm eliminated the need of training
sequence and can combat the effects of noise and interference
for high performance.It only employs the information of the
modulation type, the baseband shape and the symbol rate of
source signals to obtain cyclic autocorrelation function. A
combinatorial analysis and parameter estimation method are
presented. Some computer simulation results are demonstrated to
confirm the channel identification and subsequent equalization
methods.

並列關鍵字

Cyclostationary ； multipath ； blind ； identification ； equalization