Hybrid Subcarrier Selection Method for Vital Sign Monitoring With Long-Term and Short-Term Data Considerations

Abstract

Tracking vital signs during sleeping is important because the breathing and heart rates can be used to detect breathing diseases and heart disorders. However, existing work for breathing and heart rate estimations only considered long-term data, while ignoring the fact that people could turn over frequently in practice, which results in different data lengths of breathing and heart rate signals. In addition, most systems based on channel state information (CSI) only adopted either amplitude or phase information to select subcarrier, which could degrade the performance of breathing and heart rate estimations. In this article, in order to improve the performance, we first propose a new subcarrier selection method by jointly considering the CSI amplitude and phase information. Then, based on the proposed subcarrier selection scheme, a new breathing and heart rate estimations scheme is proposed under the consideration of different data lengths, where the peak detection and the FFT-based methods are, respectively, used for long-term and short-term breathing signals, while the FFT-based method is used to estimate the heart rate for both long-term and short-term signals. Experiment results demonstrate that our work can achieve accuracies of 96.1% and 94.3% for breathing and heart rate estimations.