A Noncontact Vital Signs Monitoring Based On Milimeter Wave Sensing

Vital sign monitoring is the key technology to ensure human health.Particularly,breath and heartbeat,as the important sign of human health,have become an important research direction.At present,vital sign monitoring is generally divided into:contact monitoring based on wearable devices and contactless monitoring based on wireless signal.The former requires the user to wear sensing devices actively and is not suitable for long-term monitoring;the latter can only achieve heartbeat monitoring within 1.5 meters due to signal attenuation and decease in directionality of beam.To solve this problem,this paper proposes a long-distance contactless vital sign monitoring system based on multi-channel millimeter wave signals.In the study of long-distance respiratory and heartbeat monitoring,we faced two main challenges when millimeter wave signals propagate over long distances:significant energy attenuation and decreased directivity.We targeted the use of multi-channel antenna array to enhance the energy of the effective signal,and adopted the minimum variance distortionless response(MVDR)adaptively to accurately locate the chest from multiple signal sources.Then,we extracted the principal components of chess motion.Finally,we calculated the breathing rate based on the phase slope and heartbeat rate based on image segmentation using CNN network.At the same time,we evaluate the accuracy of different static postures in daily life such and sitting and lying.Experimental results show that we can achieve respiratory mean error of 1BPM and heart beat mean error less than 4BPM within 2.5m.Our results show that the monitoring distance and accuracy can satisfy the demand of daily,long-term monitoring.