低阶非高斯有色噪声的Wigner-Ville谱  

Lower-order Wigner-Ville Spectrum in Non-gaussian Colored Noise Environments

作　　者：林政剑[1] 李卫升[2] 查代奉[1] 

机构地区：[1]九江学院电子工程学院 [2]九江职业技术学院基础部,江西九江332005

出　　处：《九江学院学报（自然科学版）》2010年第3期15-20,共6页Journal of Jiujiang University：Natural Science Edition

基　　金：国家自然科学基金资助课题(60772037);江西省卫生厅科技项目(20092076);江西省教育厅科技项目(GJJ10622)

摘　　要：文章对无限方差非高斯有色噪声进行时频分析,介绍了Wigner-Ville谱及分数低阶协方差的概念;定义了分数低阶Wigner-Ville谱;讨论了分数低阶Wigner-Ville谱的性质。α稳定分布的随机过程,没有有限的二阶矩,但可以利用非线性变换函数抑制稳定分布尖峰脉冲的特性,对α稳定分布随机过程作非线性变换,变换后的随机过程存在二阶矩。在有用信号中加入不同的稳定分布噪声分别进行不同的非线性变换后分析其Wign-er-Ville谱,计算机仿真实验表明,经过非线性变换后的Wigner-Ville谱能够对α稳定分布噪声背景下的信号进行有效识别,是一种有效的α稳定分布的时频域分析工具。This paper analyzed the time-frequency domain characters of non-gaussian noise with infinite variance.Wigner-Ville spectrum and fractional lower order covariance were introduced.A definition of fractional lower order Wigner-Ville spectrum was proposed and it＇s properties were discussed.Stable distribution process had not finite second-order moment.Non-linear transform functions possess the character of restraining outliers in stable distribution.A non-linear transformation is used in stable distribution process for getting finite second-order moment.The available signal in different stable distribution noise enviroment was implemented by different non-linear transformation,then it＇s Wigner-Ville spectrums were analyzed.Computer simulation results showed that Wigner-Ville spectrum which can identify useful signal effectively in stable distribution noise after a non-linear transformation was an effective tool of time-frequency analysis for stable distribution process.

关 键 词：时频分析 Α稳定分布 分数低阶协方差 非线性变换 分数低阶Wigner-Ville谱 

分 类 号：TN911.6[电子电信—通信与信息系统] TN911.7[电子电信—信息与通信工程]