Research On Sleep State Monitoring Technology Based On Channel State Information

Tracking vital signs of human sleep posture and breathing and heart rate during sleep is important because it can help assess a person’s general physical health and provide useful clues to diagnose possible diseases.Traditional methods are limited to clinical use,and RF-based methods require specialized equipment or dedicated wireless sensors and can only track breathing rates.Since human breathing and heartbeat can cause weak movements in the abdomen and chest that can have an impact on the propagation of WiFi signals,and fine-grained channel state information(Channel State Information,CSI)in WiFi devices can record these effects,this article is intended to provide a low-cost,continuous and non-contact vital sign monitoring and sleep posture recognition system based on WiFi signals.The main contributions of this work are summarized as follows:1)Because the signal will be affected by direct path in the process of transmission and reflection path of various factors such as interference,so will exist in the process of transmission attenuation of the signal,this article through the analysis of the Fresnel zone and rice distribution model,so as to put forward a method of motion detection ability enhancement are used to improve the small deformation caused by breathing and heartbeat detection ability,It is applied to the study of sleep position recognition.2)Because there are all kinds of abnormal movement during sleep,such as roll over,up to a different impact on signal,so this paper proposes a localization algorithm based on regular detection of abnormal movement,it can accurately positioning is different from breathing and heartbeat of the location of the abnormal movement,and carries on the filter,in order to obtain better monitoring vital signs.3)In this paper,through the processing of wireless signal CSI,a real-time processing system is designed to realize the real-time monitoring of respiration and heartbeat in different sleep postures.The experimental results show that the performance of this method is good,and the accuracy of respiration and heart rate detection is 96.618% and94.708%,respectively.In addition,the existing WIFI technology can also be used to track people’s sleep posture.In three typical classifiers,the average recognition accuracy reaches more than 93%.This WIFI-based sleep state monitoring method is of great significance for the monitoring of vital signs in non-clinical environments.