Publication

Abstract : This research contributes novel insights to the field of power management in real-time embedded systems, specifically through the application of dynamic voltage frequency scaling (DVFS) for CMOS-based devices. It advances beyond existing literature by critically analyzing and offering innovative solutions in power-aware scheduling, a key yet underexplored aspect in energy efficiency. The study challenges traditional approaches focused on increasing battery capacity, proposing instead a more sustainable, efficiency-centric strategy. A detailed examination of various DVFS techniques is conducted. Python-based simulations facilitate a comparative analysis against traditional algorithms like EDF and its derivatives. The findings distinctly highlight that dynamic methods are markedly more energy-efficient than their static counterparts. This increased efficiency is primarily due to their reliance on actual execution times, a significant leap from the worst-case scenario approach commonly used in static methods. By bridging a crucial gap in current research, this paper not only enriches the theoretical understanding of power-aware scheduling but also validates its practical application, marking a significant advancement in the realm of energy management for digital systems.