Research On Human Respiratory Detection System Based On WIFI

In recent years,with the improvement of life quality,people pay more and more attention to their health.As an important part of human health,sleep has been paid more attention by more people.At present,there are also many sleep-sensing devices on the market,such as smart bracelets,which can be used to detect the quality of sleep.However,these tests are all contact,which brings inconvenience to daily life and cannot be monitored for a long time.Therefore,some non-contact perceptual means are needed to monitor the human body.There is great potential for respiratory detection using channel state information(CSI)retrieved from commercial Wi Fi devices.However,the existing methods(Fresnel area detection or phase detection)are only effective when the target is close to the Wi Fi transceiver,but when the target is far away,the performance will be significantly reduced,and the perceived distance greatly limits its application in real life.This paper proposes a breath detection method of Far BR(Far Distance Breath Monitoring)based on CSI.Firstly,the ratio of CSI between the two antennas is used,and it is verified that the CSI ratio model can eliminate the random phase error caused by THE CSI hardware measurement,and can more stably use the CSI phase information that is more sensitive to the tiny motion,thus achieving a longer perceptual distance.Based on the CSI ratio,this paper further performs autocorrelation(ACF)operation on the CSI ratio power to obtain the respiratory frequency in real time after one respiratory cycle(5-7 seconds).In order to prevent the possible detection deviation of a single sub-carrier signal,the paper maximizes the signal-to-noise ratio of the power autocorrelation signal of the CSI ratio by combining multiple sub-carrier signals,and on this basis,the optimal respiration cycle is estimated.In this paper,an experimental platform is built for breath detection based on Far BR algorithm.Experiments were carried out at different distances from the receiver and in different postures.The experimental results show that the scheme based on Far BR algorithm has a long perceptual distance and accurate respiration rate estimation.