基于循环平稳特性的两步盲信号提取算法  

Cyclostationarity-based Two-stage Blind Signals Extraction Algorithm

作　　者：尚晓辉[1] 沈越泓[2] 王建功[3] 黄福贵 

机构地区：[1]解放军理工大学通信工程学院研究生1队,江苏南京210007 [2]解放军理工大学通信工程学院无线通信系 [3]解放军理工大学通信工程学院研究生2队 [4]海军驻景德镇地区航空军事代表室,江西景德镇333000

出　　处：《军事通信技术》2011年第4期62-66,共5页Journal of Military Communications Technology

摘　　要：盲提取算法作为信号处理和分析的重要手段,其利用信号的先验特性提取期望信号,从而降低算法计算量、节约系统资源。文章针对通信信号的盲提取问题,基于信号的循环平稳特性,提出了一种两步盲提取算法:第一步,利用信号的循环平稳先验特性粗提取信号;第二步利用信号的高阶统计特性精提取信号。理论分析与计算机仿真表明:与参考文献[5]中基于循环平稳特性的盲提取算法相比,提出算法具有更好的性能,平均信噪比可提高约10 dB。As an important means of signal processing and analysis,blind extraction algorithm only seeks to extract ＂interesting＂ source signals with prior information so as to save computing time and system resources.To effectively extract communication signals,a two-stage blind extraction algorithm based on cyclostationarity of the original sources was proposed.At the first stage,the cyclostationary property was exploited to roughly extract the desired source signal. At the second stage,the extract signal was refined based on the higher-order statistic.Theoretical analysis and computer simulation show that compared with the algorithm of literature [5]based on the cyclostationary property,this algorithm can achieve a better performance with an average SNR improvement by about 10 dB.

关 键 词：盲源提取 循环平稳 一元固定点算法 

分 类 号：TN911.23[电子电信—通信与信息系统]