The sleep disorders concern many people in the advanced societies, so the development of the devices and methods of sleep monitoring became a very important one. The examination of the snore plays a big role in the course of sleep examination, because the reason for the lack of sleep may be a considerable one (mainly its single types, for example sleep apnoe syndrome). "The loud snoring with breathing pauses are associated with cardiovascular disorders and increased health-care utilization." - according to a newspaper SLEEP 2008th article published in the March issue . Because of the personality rights it is not permitted to record the sounds with microphones, then send or store them. So, it appears expedient to process the sound locally without storing, and to send or store just the information about snoring.
During the work, a hardware has been developed, which uses a microphone to record the sounds generated during sleep. It is able to process these independently and to recognize the fact of snoring. The recorded sound is not forwarded to other units, and are not saved thus avoiding the inherent potential privacy problems. Another important requirement is that the observed person should be disturbed as little as possible during sleep. Course of the so-called polysomnographic trial, a person must spend a night in a sleep laboratory, and has to sleep with all sort of sensors and electronic observers attached her body. By contrast, my goal is to be able to observe the patient at her own home, without body- attached sensors.
The goal of my paper is to present a snore detection solution based on this hardware device, despite having limited resources. At first I present the signal processing algorithms being used for detecting the snore [2,3], and then I’m going to analyse it in terms of the real-time reliability. In the second step I present the realisation and testing of the selected algorithms based on the microcontroller environment.
1. A Dunai, A Keszei, MS Kopp, CM Shapiro, I Mucsi, M Novak, "Cardiovascular disease and health care utilization in snorers: a population survey" SLEEP, Vol 31/Issue 3 March 1, 2008
2. Li Tan and Montri Karnjanadecha, "Pitch detection algorithm: Autocorrelation Method and AMDF" Proceedings of the 3rd International Symposium on Communications and Information Technology, 2: 551-556, September 2003.
3. M.R. Schroeder, "Period histogram and product spectrum: New methods for fundamental frequency measurement" JASA, 43(4):829-834, 1968.