The document discusses using wavelet transform to analyze snoring signals. Wavelet transform is proposed as an adequate method for analyzing non-stationary snoring signals because it allows a focus on spectral characteristics over time. Experimental results showed that wavelet transform has advantages in snoring signal analysis and processing, as it can easily and effectively detect signals containing different components. Wavelet transform can decompose mixed signals into different frequency components and sample frequencies of different sizes in the time-space domain, allowing comprehensive analysis of varying signal spectra.

1. The Research of Wavelet Transform for the Snoring Signal Zhang Yinhong ， Li Quanlu College of Physics and Information Technology, Shaanxi Normal University, Xi’an, Shaanxi 710062, China

2. Introduction Obstructive sleep apnea (OSA) is a highly prevalent disease in which upper airways are collapsed during sleep, leading to serious consequences . The OSAS has important clinical implications, ranging from disruption of sleep with daytime sequelae of excessive sleepiness to suspected cardiovascular consequences in the long term. Snoring is a typical non-stationary signal, wavelet transform is proposed as a more adequate method because it gives the possibility to focus on the signal spectral characteristics. For the characteristics of snoring signals, it analyzes the characteristics of each spectrum, timely and accurate diagnosis of respiratory obstruction caused by snoring sound pathological signals, and keeps abreast of the upper respiratory tract the extent of obstruction, understands the potential dysfunction.

4. Results And Discussion

5. Results And Discussion Experimental results show that wavelet transform has obvious advantages in the snore sound signal analysis and processing and that can detect easily and effectively the signal containing different components. The greater degree of obstruction, the more narrow band of the signal, its spectrum of low-frequency part would be filtered out, vice versa. Wavelet transform can have a variety of different frequencies intertwined with the composition of mixed-signal decomposed into different frequency signals, and it can decomposed mixed signal which include various interwoven with different frequency components into different frequency signals, and stratified sampling the frequency of different size with corresponding signal in time-space domain, thus it can step into any objects comprehensive details and vary signal spectrum for analysising.

6. Thank you 2010-03