Respiration and Heartbeat Rates Measurement Based on Convolutional Sparse Coding

Accurate access to respiration rate (RR) and heartbeat rate (HR) through radar is of great importance in many applications. In this paper, a novel method based on convolutional sparse coding (CSC) is proposed for respiration and heartbeat rates measurement. To solve the problem of algorithm performance degradation caused by insufficient samples, a number of samples are generated by mixing random noise with original signal. Then radar signals are processed by CSC directly in the time domain. The method is tested by a vital sign data generated by finite differences time domain (FDTD) simulation. The results demonstrate that the proposed processing approach can accurately extract the respiration and heartbeat components with the generated data of 5 seconds.