Research And Implementation Of Respiratory Rate Detection System Based On WiFi Signa

At present,more than 1 billion people in the world have respiratory diseases,causing more than 4.2 million deaths.The home respiratory rate monitoring system can detect the respiratory rate of a person and evaluate the respiratory function.The non-contact respiratory monitoring system based on WiFi has received wide attention from scholars because of its non-invasive and low-cost advantages.However,the existing WiFi respiratory monitoring system lacks in-depth analysis of the detection boundary of the respiratory frequency,and the detection boundary is of great significance to the non-contact detection system,which can reduce the potential radiation risk and improve the availability of the monitoring system.This paper proposes a WiFi-based respiratory frequency detection boundary estimation model,which can provide an important theoretical basis and reference for the evaluation of the detectable ability of the respiratory monitoring system.The main work of this paper includes the following three aspects.(1)Under the condition of detectable breathing frequency,in order to accurately estimate the breathing frequency detection boundary in the real sleep scene,this paper studies and implements a WiFi-based breathing frequency detection boundary estimation system.This article first analyzes the impact of human breathing behavior on WiFi wireless channels,and then derives a multipath wireless channel model that is interfered by breathing,reveals the relationship between CSI amplitude and breathing frequency,and gives the detection boundary conditions of the judgment theory.Finally,combined with the theoretical detection boundary conditions,a detection boundary system for estimating the actual respiratory frequency is designed.(2)In order to achieve accurate respiratory frequency estimation results when the WiFi device is deployed as far as possible from the sleeper within the respiratory frequency detection boundary,this paper studies and implements a respiration based on wavelet transform filtering and sub-carrier selection Frequency detection system.The system is mainly composed of four parts,namely,outlier filtering,wavelet transform filtering noise,sub-carrier selection and FFT estimation of respiratory frequency.First,the hampel filter is used to remove the abnormal values ??of the original CSI amplitude data,and then the wavelet transform is used to decompose and reconstruct the CSI amplitude sequence to filter out noise and reduce the interference of noise on the respiratory signal;then select the optimal subcarrier to obtain the corresponding CSI amplitude sequence,and finally use FFT to obtain the estimated breathing frequency of the sleeper.(3)Through simulation and experiments in the upper and lower berth scene and the low height bed scene,it is found that the actual detection boundary value is close to the theoretical detection boundary value,which verifies the validity and correctness of the respiratory frequency detection boundary model derived in this paper.Compared with the two existing respiratory frequency detection systems,the respiratory frequency estimation system proposed in this paper can still obtain accurate detection results when the device is near the detection boundary,and the relative error of the respiratory frequency estimation is reduced by 13% and 10%,respectively.In addition,this article also analyzes the effect of the distance between the transmitting and receiving antennas and different sleeping positions on the performance of the respiratory frequency detection system.When the sleeper is lying on his side and the receiver is placed far away from the sleeper,the estimated relative error of the respiratory rate of the respiratory rate detection system designed in this paper is still only 11%.