Energy-Efficient Ultra-Low Power PPG Wearable System for Long Term IoT Health Monitoring

This work presents an ultra-low power (ULP) photoplethysmography (PPG) end-to-end wearable system that offers continuous physiological monitoring capabilities for Internet-of-Things (IoT) applications. The increasing demand for wearable health monitoring devices necessitates the development of such energy-efficient PPG systems. Leveraging existing ULP System-on-Chip (SoC), Analog-Front End (AFE) chip, and a Bluetooth Low-Energy Transmitter (BLE-TX) chip, our work integrates these chips into an integrated wearable IoT system for continuous real-time PPG health-monitoring. Employing a systems-level driven approach, rapid prototyping of the system leveraging the flexible design of the ULP chips enables a ‘plug-and-play’ approach towards IoT system design. All components are fabricated using the TSMC 65nm Low-Power (LP) CMOS process. This system achieves minimal power consumption at just 62.8 µW during idle periods and 148.5 µW during active data transmission when aggressively duty-cycled at 2.5%. This compact system measures in at just 75mm x 70mm x 35mm and achieves an operational period of 147 days when powered by a high-capacity 3.7V, 500-mAh CR3555 rechargeable battery, making it suitable for ULP IoT applications. This thesis will first by introduce a modular design language for low-powered end-to-end systems, followed by a practical demonstration of this concept through PPG system integration, from the process of PCB design, system testing, to optimization for low-power.