Improved Multi-Person Vital Signs Estimation Algorithm in Sitting and Standing Positions using MIMO FMCW Radar

The use of FMCW radars for non-contact monitoring of vital signs has gained an increased attention in the healthcare sector within the past few years. In this paper, we propose an improved contactless vital signs estimation algorithm that is capable of detecting multiple persons, both in sitting and standing positions, and accurately estimating their respiratory and heart rates. The proposed algorithm combines several state-of-the-art radar vital signs estimation approaches, including empirical mode decomposition with principal component analysis, density-based clustering, and Kalman filtering. The proposed algorithm was developed and tested using a market-available FMCW radar in MIMO configuration. The testing results showed that the algorithm can estimate the respiratory and heart rates with an error of 0.7 cycles/minute and 2.3 beats/minute, respectively.