효율적인 신호분리기법을 이용한 근접한 두 사람들의 비접촉식 생체신호 탐지

Abstract

Recently, non-contact vital sign(i.e. respiratory and cardiac rate) detection(VSD) techniques based on radar sensors have been developed for several applications in the medical, Internet of Things(IoT), and military fields. However, if two humans are located at the same distance, the VSD of these methods may fail. This is because the two humans yield mixed and complicated echo signals, thereby making the VSD task quite difficult. To overcome this problem, several signal decomposition methods, such as variational mode decomposition and multiple signal classification, have been developed. Nevertheless, mixed echo signals from two humans at the same distance cannot be decomposed. In this study, we propose a new signal decomposition method to achieve the VSD of humans at the same distance. In particular, the proposed method can simply decompose I- and Q-channel parts of a mixed echo signal into the individual phase fluctuations corresponding to each human, thereby providing improved computational efficiency. In experiments using the impulse-radio ultra-wideband radar, it was observed that the proposed algorithm can successfully achieve the VSD of two humans at the same distance.