Exploiting Dynamic Phase Information for Respiration Monitoring During Sleep via WiFi

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wireless Sensing

Background:

• Respiration monitoring is crucial for diagnosing sleep disorders.
• WiFi sensing offers a non-invasive, comfortable alternative to wearable devices.
• Static components in WiFi signals hinder accurate respiration detection.

Purpose of the Study:

• To develop a robust method for extracting respiration signals from WiFi Channel State Information (CSI).
• To overcome limitations posed by static interference in WiFi-based respiration sensing.
• To improve the accuracy of respiration rate estimation for sleep-related disease diagnosis.

Main Methods:

• Exploiting dynamic phase information within WiFi CSI.
• Utilizing nonlinear least squares fitting to eliminate static signal components.
• Applying Principal Component Analysis (PCA) to extract dominant respiration signals and reduce noise.

Main Results:

• The proposed method effectively extracts respiration signals from WiFi CSI.
• Accurate respiration rate estimation was achieved across various sleep postures.
• The approach demonstrated superior performance compared to existing state-of-the-art methods.

Conclusions:

• The developed WiFi sensing technique provides a promising, non-contact solution for sleep respiration monitoring.
• This method significantly enhances the accuracy of respiration rate estimation, aiding in sleep disorder diagnosis.
• The approach offers a comfortable and convenient alternative to traditional wearable sensors.