A Novel Period Estimation Method for Periodic Signals Suitable for Vital Sensing

The Internet of Things (IoT) has been widely used in care facilities such as hospitals and nursing homes. These facilities make efficiently health management of its patient or the elderly by monitoring vital signs using the IoT. The non-contact vital sensing is particularly interesting as a measuring method for monitoring since it does not need electrodes. It enables us to release patients from constraints by cables and electrodes. The Doppler sensor is a mean of the non-contact vital sensing. We can detect respirations and heart beats by the electro-magnetic wave. Considering the non-contact vital sensing using the Doppler sensor, there are two problems. First, vital signs obtained by the Doppler sensor is often noisy. Therefore, the signal processing algorithm must be robust against noisy vital signs. Second, the frequency of vital signs is typically very low such like less than 1 Hz. It is well-known that the fast Fourier transform (FFT) suffer from low resolution in such very low frequency band. In this paper, we propose a new signal parameter estimation algorithm which is robust against the noise while it achieves high resolution in the low frequency band. The performance of the proposed method is provided in this paper.