Breath Rate Detection in Single and Multi-User Scenarios Using Wi-Fi Channel State Information

Abstract

This study presents a signal processing pipeline for non-intrusive breath rate detection in single-user and multiuser scenarios using wireless Channel State Information (CSI). Our approach integrates advanced phase cleaning, Butterworth filtering, and FFT-based feature extraction to accurately estimate respiratory rates in single-user and multi-user scenarios. Experimental results show competitive performance with a Mean Squared Error (MSE) as low as 0.50 BPM in controlled settings. We also discuss the challenges in separating overlapping signals in multi-user environments and propose future work incorporating adaptive filtering and deep learning models for real-time monitoring.