一种同频混合信号伪随机序列盲估计方法的研究  

Research on a pseudorandom sequence blind estimation method of co-frequency mixed signal

作　　者：李在林[1] 王松波 LI Zailin;WANG Songbo(College of Physics and Electronic Engineering,Xinxiang University,Xinxiang 453000,China;Department of Computer Science,Zhengzhou Economy and Trade School,Zhengzhou 450053,China)

机构地区：[1]新乡学院物理与电子工程学院,河南新乡453000 [2]郑州市经济贸易学院计算机系,河南郑州450053

出　　处：《现代电子技术》2017年第9期10-13,共4页Modern Electronics Technique

摘　　要：通常采用独立分量分析法(ICA)采集同频混合信号时存在盲分离随机性问题,不能分离出同频混合信号伪随机序列,无法对信号进行准确检测。为解决该问题,提出融合独立分量分析法以及Massye算法的同频混合信号伪随机序列盲估计方法。先采集同频混合信号,再通过PCA方法对同频混合信号进行白化预处理,对同频混合信号的协方差矩阵的特征值进行分解,确保信号间相互独立,为后续ICA方法进行数据分割提供基础。采用基于峰度的固定点ICA算法对白化处理后的同频混合数据进行划分,融合ICA和Massye算法,对同频混合信号的伪随机序列进行盲估计。实验结果说明,该方法可以获取准确的同频混合信号伪随机序列,具有较强的信号分离性能。The independent component analysis(ICA)method is usually used to collect the cofrequency mixed signal,but it has the problem of blind separation randomness,and it can′t separate the pseudorandom sequence of the cofrequency mixed signal,nor can it detect the signal accurately.In order to solve the above problems,a cofrequency mixed signal′s pseudorandom sequence blind estimation method based on ICA method and Massye algorithm is proposed.The cofrequency mixed signalis acquired,and performed with the whitening pretreatment by means of PCA method.The eigenvalue of the covariance matrix of the cofrequency mixed signal is decomposed to ensure the mutual independence among the signals,and provide the basis fordata segmentation with the follow-up ICA method.The fixedpoint ICA algorithm based on kurtosis is used to divide the co-fre-quency mixed data after whitening treatment.The ICA algorithm and Massye algorithm are fused to perform the blind estimationto the pseudorandom sequence of the cofrequency mixed signal.The experimental results show that the proposed method can obtain the accurate pseudorandom sequence of the cofrequency mixed signal,and has strong signal separation performance.

关 键 词：同频混合信号 伪随机序列 盲估计 Massye算法 

分 类 号：TN911-34[电子电信—通信与信息系统]