Distortion Reduction and Signal Estimation in Doppler Radar Physiological Monitoring Systems

Abstract

Feasibility of microwave Doppler radar for non-contact physiological monitoring and detection is explored in this dissertation. Although there was significant progress during the last decade in this technology, there are still numerous challenges to be addressed. In this work several sources of signal distortion are investigated. Contributions to the field of electrical engineering include novel hardware and software techniques to overcome distortions and improve signal estimation in wireless vital sign monitoring. Channel imbalance in quadrature receivers is one of the distortion sources that has been systematically investigated, theoretically and experimentally. It has been demonstrated that imbalance can introduce error in displacement estimation, radar cross section, and even rate measurements. Innovative hardware approaches, including packet radar and pulse low-IF receiver architectures, were proposed to resolve AC coupling issues, channel imbalance, and position sensitivity. Signal estimation for single and multiple subject detection was explored. True human presence estimation is demonstrated with Doppler radar occupancy sensors based on heart and respiration rates to overcome the limitations of common occupancy sensors. Estimating direction of arrival in a single input multiple output system is explored for multiple subject detection. These new approaches may bring Doppler radar physiological monitoring devices one step closer for reliable system performance at low power and low cost.