Heart Rate Variability and Body-Movement Extractions using Wi-Fi Channel State Information

Sleep is an important physiological process and is known to be closely related to human health. Polysomnography is a common method for analyzing sleep cycles. However, this method is not suitable for routine monitoring because it requires that the patient wears multiple sensors and the results need to be interpreted by a physician. The channel state information obtained from commodity Wi-Fi devices can be used to analyze complex multipaths and enables noncontact measurement of biometric information. In this study, we propose a non-contact extraction system of heart rate variability and body-movements as datasets for classifying sleep stages. The heart rate variability analysis, which uses variational mode decomposition, shows an average accuracy of 88.9 ms within 1 m between the transmitter and receiver. In addition, the proposed body-movements analysis system was shown to provide highly accurate estimation in real-time.