Article Sidebar
Main Article Content
Abstract
Smart healthcare monitoring systems have become increasingly important for continuous patient observation and timely medical intervention. However, many low-cost embedded healthcare devices are developed with limited consideration of real-time requirements, which may lead to delayed responses during emergency situations. This paper presents a real-time requirements analysis of an ATmega328P-based healthcare monitoring device capable of measuring body temperature, heart rate, blood oxygen saturation (SpO₂), and blood pressure while providing remote alert functionality through GSM communication.
The study begins with an analysis of the implemented prototype and its software architecture. The original system employs a sequential super-loop execution model containing multiple blocking delays that negatively affect responsiveness and emergency handling performance. To address these limitations, a hybrid real-time architecture is proposed that combines event-driven processing, interrupt-assisted emergency detection, and priority-based task scheduling.
Worst-Case Execution Time (WCET) estimation, response-time analysis, Rate Monotonic Scheduling (RMS), and Earliest Deadline First (EDF) scheduling techniques were applied to evaluate the timing behavior of the system. A Python-based simulation framework was developed to assess processor utilization, schedulability, response times, deadline satisfaction, and emergency alert latency.
The results show that the proposed task set achieves a processor utilization of 14.5% and remains schedulable under both RMS and EDF policies. Response-time analysis confirmed that all tasks meet their timing constraints. Furthermore, the proposed hybrid architecture reduced emergency alert latency from approximately 20.5 seconds in the original implementation to 160 milliseconds, representing a latency reduction of 99.22%.
The findings demonstrate that integrating real-time scheduling mechanisms and interrupt-assisted emergency management can significantly improve the responsiveness, predictability, and reliability of resource-constrained healthcare monitoring systems while maintaining compatibility with low-cost embedded platforms.