一种基于小波包变换的双模噪声中信号检测  

Signal Detection in Background of Bimodal Noise Based on Wavelet Packet Transform

作　　者：孙万麟[1] 山拜.达拉拜 

机构地区：[1]昌吉学院,新疆昌吉831100

出　　处：《现代电子技术》2009年第13期63-65,共3页Modern Electronics Technique

基　　金：国家自然科学基金资助项目(60062001)

摘　　要：利用多尺度小波包变换良好的时频局部分析能力,对一类非高斯噪声——双模噪声的统计特性进行研究。经过实际计算及仿真比较,对于两种双模噪声模型,随着双模噪声序列长度的增加,双模噪声小波包子空间变换系数都将近似服从高斯分布。通过仿真,证实小波包检测系统的检测性能不仅跟输入双模噪声的模型无关,而且比传统的高阶统计量检测系统、经典检测系统的检测性能都要好。On the basis of time- frequency analysis capabilities of multi- scale wavelet packet transform, statistics of one special of non- Gaussian noise- bimodal noise are researched. Through the actual calculation and simulation, with the increase of its length of the sequence,its subspace wavelet packet transform coefficients are all similar to obey Gaussian distribution for the two bimodal noise model. The simulation not only confirms that its detection performance has nothing to do with the bimodal noise model, but also its detection performance is better than the traditional higher -order statistics detection system and the classical detection system.

关 键 词：小波包变换 双模噪声 检测性能 统计特性 

分 类 号：TN911[电子电信—通信与信息系统]