A Robust Respiration Detection System Via Similarity-Based Selection Mechanism Using WiFi.

Xinyi Zhou,Ting Jiang,Xue Ding,Sai Zhang,Yi Zhong

DOI: https://doi.org/10.1109/wcnc55385.2023.10118897

2023-01-01

Abstract:Recent research has demonstrated the great potential of leveraging existing WiFi infrastructure for ubiquitous non-invasive respiration monitoring. Although this WiFi-based approach opens up a new direction for respiratory rate detection, existing studies are limited as only some simple scenarios have been considered. Consequently, the feasibility of using this technology in realistic scenarios needs to be further verified, especially for ensuring the detection performance in the following two cases: (1) long-distance and (2) different body postures. To address above two complex case studies, this paper presents several selection mechanisms to enable a robust WiFi-based respiration detection system. Firstly, a double-variance antenna links selection strategy is proposed to select the most sensitive link for breathing movements. Moreover, three subcarrier selection combining solutions are developed, where secondary selection is conducted to obtain the optimal respiration pattern in diverse situations. We conduct extensive experiments in two typical scenes. The evaluation results demonstrate that the detection error of our system is less than 0.7 bpm in each scene. More importantly, it outperforms compared with state-of-the-art systems.