Intrinsic Respiration Calibration Technique Based on EMD to Enhance Accuracy of Heartbeat Signal Contrast in a Hybrid FMCW-CW Radar System

Abstract

This thesis presents an empirical model decomposition (EMD)-based method to enhance SNR of heartbeat signal obtained from a hybrid FMCW-CW 5 8 GHz ISM band radar system EMD can solve the problem of being hindered by higher order of respiration for finding spectra of heart rate (HR) Certain intrinsic mode functions (IMFs) based on content of power ratio would be combined to rebuild respiration signal in time domain Heartbeat signal from both FMCW and CW modes were extracted after respiration signal calibration Experiment scenarios including target of healthy human and metal plate were carried out to detect vital sign as well as absolute range information respectively The measurement results suggested that the extracted heartbeat signal was significantly improved by 40 74 dB from CW mode and by 24 1 dB from FMCW mode in SNR The spectra of extracted HR indicated accuracy of heartbeat signal within error <1 % These findings have implications for hybrid radar system locating and acquiring physiological information simultaneously

Date of Award	2017 Aug 22
Original language	English
Supervisor	Chin-Lung Yang (Supervisor)