| dc.description.abstract | The growing global demand for unobtrusive, non-contact health monitoring systems, driven by aging populations and the increasing prevalence of neurological disorders, underscores the urgent need for innovative sensing technologies that seamlessly integrate into everyday indoor environments. In response, this research advances portable Frequency Modulated Continuous Wave (FMCW) radar systems as a promising solution for continuous biomedical motion sensing, combining rigorous simulation-led design with novel hardware and waveform innovations.
A high-fidelity three-dimensional simulation framework was first established to systematically characterize human motion signatures, offering critical insights into range Doppler performance under diverse radar configurations, environmental clutter, and system non-idealities. By simulating multiple frequency bands and carefully modeling
multipath propagation and phase noise, the framework enabled precise evaluation of trade -offs in range resolution, velocity discrimination, and clutter resilience, establishing an essential foundation for subsequent hardware development.
Building on these insights, the thesis details the realization of a compact radar platform that integrates a high-gain quasi
-Yagi antenna with an S-band radar-on-chip (RoC) system. Extensive bench-top experiments and through-wall detection trials demonstrated the platform’s capability to reliably detect both static structures and dynamic human motions across a variety of challenging indoor environments. This practical validation affirmed the viability of deploying such portable radar systems for in-home monitoring, where privacy preservation and continuous operation are paramount.
Recognizing the stringent phase coherence and synchronization demands of conventional FMCW architectures for accurate Doppler extraction, this research proposed and validated a novel hybrid waveform. By embedding continuous wave segments within each FMCW chirp, the approach fundamentally relaxes hardware constraints, eliminating external clock synchronization while maintaining phase integrity.
Rigorous MATLAB simulations and extensive indoor experiments confirmed its ability to deliver clean range-Doppler estimates, even amid clutter, multipath, and electromagnetic interference. Finally, the system was validated in scenarios capturing vital sign motions (VSM) and upper-limb tremors, VSM being pertinent to general physiological monitoring, while tremor quantification holds diagnostic value in early neurological assessments. These preliminary results highlight the radar’s sensitivity to micro-movements and its robustness in non-contact measurement settings. This lays a strong foundation for future clinical trials across diverse participant cohorts and real-world deployment scenarios, ultimately aiming to validate the system’s utility in continuous, in-home monitoring and early-stage disease detection frameworks.
Collectively, this thesis delivers a unified framework spanning simulation-led radar optimization, portable hardware innovation, and hybrid waveform design, advancing FMCW radar toward practical, widespread adoption in biomedical monitoring. Future work will explore adaptive signal processing and AI-driven waveform selection, evolving the system into a cognitive radar platform that autonomously adapts to dynamic environments, with applications extending to smart infrastructure, automotive safety, and advanced surveillance. | en_US |
