Deep Neural Network based Body Movement Cancellation for Doppler Radar Vital Sign Detection

This paper presents a novel technique of cancelling random body movement (BM) in Doppler radar noncontact vital sign detection based on deep neural network (DNN). Respiration motions with different frequencies and different phase offsets are synthesized and mixed with simulated BMs to train the DNN model. Neither hardware modification nor extra hardware complexity is needed. The proposed DNN based method can separate the stable respiration motion from the signal mixture of respiration and BM without prior information from the BM. Experimental results show that the proposed DNN based method allows a single radar sensor to successfully reconstruct stable respiration motions without BM disturbance.