Noncontact healthy status sensing using low-power digital-IF Doppler radar

Abstract

Health status sensing is of great significance in early disease prevention, clinical treatment and back-end home care. Vital signs can not only provide physiological information but also reflect various health statuses of human subjects. Our emphasis is on discovering inner relationships between the Doppler radar-based noncontact vital sign detection and the health status of human subjects. The custom-designed low-power digital intermediate frequency (digital-IF) continuous-wave (CW) Doppler radar has been designed to capture vital signs with high accuracy and robustness. Then the compressed sensing (CS)-based method is proposed to enhance the resolution of the vital sign spectrum, and the synchrosqueezing transform (SST)-based algorithm is used to extract the instantaneous vital signs. Based on the digital-IF Doppler radar and the advanced signal processing algorithms, several health status sensing modules, including the breathing disorder recognition and sleep-stage estimation, have been realized by using advanced machine learning techniques. Laboratory and clinical experiments demonstrate the effectiveness of noncontact health status sensing using the proposed radar system.