摘要
将小波包变换良好的时频分析能力用于分析双模噪声的统计特性,因为小波包变换具有带通滤波的作用,当分解的层数足够多时,变换可看作窄带系统,将双模噪声作某一尺的小波包变换,在此空间上双模噪声的输出近似于高斯分布,在这一空间上进行信号的识别与检测。将小波包变换用于双模噪声背景下的信号检测系统,并将此方法与经典检测系统进行性能上的比较,仿真结果表明,小波包方法优于经典检测方法。
A good time-frequency analysis of wavelet-packet transform is used to analyze the statistical properties of dual-mode noise. For the wavelet-packet transform has band-pass filtering effect, when the decomposition layers are sufficient enough, the change could be regarded as narrow-band system. The dual-mode noise is made a foot-long wavelet-packet transform, the output of the dual-mode noise in this space is approximated to the Gaussian distribution, and the signal is identified and detected. The wavelet-packet transform is used in signal detection system under dual-mode noise background. And the comparison of this method with classical detection system for their performance is done, the simulation result show that the wavelet-packet method is superior to the classical detection method.
出处
《通信技术》
2012年第2期125-128,共4页
Communications Technology
基金
国家自然基金科学项目(批准号:60971130)
关键词
双模噪声
小波包变换
经典检测系统
Bimode noise
wavelet-packet transforms
classical detection system
