Healthcare automation technologies have led to develop “smart” home with intelligent sensing systems to fully monitor vital signs. One important healthcare application of the millimeter-wave radar sensor is the respiratory rate (RR) estimation. Accurate and robust measurements of respiration motion and RR estimation in nonhospital facilities is a challenging task. We present a novel contactless on-bed radar system for respiration monitoring for “smart” home. A simplified trunk model and a mode-decomposition-based respiration reconstruction are designed for analysis of 2-D radar profile to extract nonstationary breathing motion information. Resting breathing experiments with different respiratory patterns and human poses are performed to evaluate the performance of the proposed system. The results show that the system measures respiration-induced body movement with a root-mean-squared error of 0.31 mm and a mean accuracy of 89.6%. For RR estimation, the overall accuracy is 97.1% under normal respiratory patterns and 93.4% under abnormal respiratory patterns.

Noncontact vital sign monitoring based on radar has attracted great interest in many fields. Heart Rate Variability (HRV), which measures the fluctuation of heartbeat intervals, has been considered as an important indicator for general health evaluation. This paper proposes a new algorithm for HRV monitoring in which frequency-modulated continuous-wave (FMCW) radar is used to separate echo signals from different distances, and the beamforming technique is adopted to improve signal quality. After the phase reflecting the chest wall motion is demodulated, the acceleration is calculated to enhance the heartbeat and suppress the impact of respiration. The time interval of each heartbeat is estimated based on the smoothed acceleration waveform. Finally, a joint optimization algorithm was developed and is used to precisely segment the acceleration signal for analyzing HRV. Experimental results from 10 participants show the potential of the proposed algorithm for obtaining a noncontact HRV estimation with high accuracy. The proposed algorithm can measure the interbeat interval (IBI) with a root mean square error (RMSE) of 14.9 ms and accurately estimate HRV parameters with an RMSE of 3.24 ms for MEAN (the average value of the IBI), 4.91 ms for the standard deviation of normal to normal (SDNN), and 9.10 ms for the root mean square of successive differences (RMSSD). These results demonstrate the effectiveness and feasibility of the proposed method in emotion recognition, sleep monitoring, and heart disease diagnosis.