基于单通道信号及NLR算法的OFDM系统频偏估计  

OFDM carrier frequency offset estimation for single channel recordings using NLR algorithm

作　　者：陈朝阳[1] 程贵宾 陈彭[1] 邢海涛[1] 

机构地区：[1]集美大学信息工程学院,厦门361021

出　　处：《仪器仪表学报》2014年第12期2852-2857,共6页Chinese Journal of Scientific Instrument

基　　金：福建省科技计划重点项目(2013H0035);厦门市科技计划项目(3502Z20123028)资助

摘　　要：提出了一种新的OFDM系统频偏估计方法,该方法首先利用相关检测技术从单路的接收信号中获得独立分量分析(ICA)技术所需的多路混合信号,相比于传统的估计方法减少了接收通道;然后使用非参数似然比(NRL)算法来实现对混合信号的分离,同时利用Hilbert变换估计OFDM系统的载波频偏(CFO)。NLR算法不需要假设信号的非高斯统计特性,且与MMI-ICA相比,NLR-ICA可以显著改善分离信号的信号干扰比(SIR)。与传统的频偏估计方法相比较,该方法不需要借助同步训练符号,不仅节省了信道带宽,而且克服了信道参数对CFO估计的影响,因此具有一定的优越性。仿真结果表明,该方法具有良好的特性。NLR算法的实现不依赖于信号的概率分布,克服了传统ICA算法对信号分布要求,有利于对通信信号的分离。A novel carrier frequency offset （CFO） estimation method is presented using the only channel recordings in OFDM systems. Firstly, multiple mixed signals are obtained from the single channel signal with the aid of the correlation detection tech-nology, which reduces receiving channels. Then the mixed signals are separated by NLR algorithm, and the CFO is estimated through Hilbert transform. Compared with MMI-ICA, NLR-ICA can significantly improve SIR of the separated signals without assuming the statistical characteristics of signals. Relying on the single channel recordings without training symbols, the method not only saves channel bandwidth, but also eliminates the channel parameter’s impact on CFO estimation. Simulation results show that the method has good properties. NLR algorithm does not depend on probability distribution of the signal, and it is good to separate communication signals.

关 键 词：单通道信号 载波频率偏移 独立分量分析 非参数似然参数比 

分 类 号：TN911.7[电子电信—通信与信息系统]