No abstract

Several studies have been published on noncontact heartbeat detection recently. Particularly, a few studies have found that radar for noncontact detection is effective for long-term monitoring. As the novel coronavirus continues to spread worldwide, the use of radar is expected to prevent the contamination of equipment by the virus and enable heartbeat detection from a safe distance. To monitor heartbeat fluctuation, heartbeat signals must be detected over a short period; however, conventional methods require approximately 5 s for detection in the stationary state of the human body. This study developed a method for detecting heartbeat in approximately 2.5 s based on nonharmonic analysis (NHA) to suppress sidelobes. The method has nearly no data length limitations in terms of frequency resolution. To validate the method, heartbeat detection is performed in a stationary state with reduced noise and a driving simulation environment with high noise. Frequency analysis of the received signals enabled the detection of heartbeat harmonics, suggesting the possibility of detecting the fundamental heartbeat frequency from its harmonics. To test this aspect, the detectability of heartbeat harmonics is determined under the stationary and driving simulation conditions from a comparison of the conventional short-time Fourier transform, continuous wavelet transforms, and the proposed NHA method.INDEX TERMS Continuous wave Doppler radar (CW Doppler radar), driving simulation, harmonic, heartbeat, non-harmonic analysis (NHA).