Abstract
Advances in micro technology has led to the development of miniaturized sensor nodes with wireless communication to perform several real-time computations. These systems are deployed wherever it is not possible to maintain a wired network infrastructure and to recharge/replace batteries and the goal is then to prolong as much as possible the lifetime of the system. In our work, we aim to modify the Earliest Deadline First (EDF) scheduling algorithm to minimize the energy consumption using the Dynamic Voltage and Frequency Selection. To this end, we propose an Energy Saving EDF (ES-EDF) algorithm that is capable of stretching the worst case execution time of tasks as much as possible without violating deadlines. We prove that ES-EDF is optimal in minimizing processor energy consumption and maximum lateness for which an upper bound on the processor energy saving is derived. In order to demonstrate the benefits of our algorithm, we evaluate it by means of simulation. Experimental results show that ES-EDF outperforms EDF and Enhanced EDF (E-EDF) algorithms in terms of both percentage of feasible task sets and energy savings.