International Journal of Computer Applications (0975 – 8887) Volume 56– No.18, October 2012 16 Real-Time Scheduling with DVS and Harvesting Energy Constraints Mona Kumari Assistant Professor Department of Information Technology Anand Engineering College, Agra Ajitesh Kumar Assistant Professor Department of Information Technology Hindustan Institute of Technology and Managenment, Agra ABSTRACT In real-time embedded system, that must carry their own power source and cannot depends on the power outlet on the wall, apart from feasibly schedule the set of tasks, power management is also the major issue because without power the system is useless. In this paper, we propose a harvesting aware real-time scheduling algorithm with variable speed assignment scheme to set of periodic tasks aims to reduce the energy consumption while feasibly schedule the set of periodic tasks within their deadline. This can be done by DVS(Dynamic Voltage and frequency Selection), executing the task with the speed such that a task can consume as much energy which is quite sufficient to complete it successfully within its deadline. The example and simulation results shows that the propose approach is capable of performing better in terms of average stored remaining energy of the system as well as acceptance ratio of periodic tasks at lower periodic load. Key Terms: Real-Time Scheduling, DVS, Energy Harvesting, Periodic Tasks, Embedded System, Power Management. 1. INTRODUCTION In real-time system to function correctly, the system must produce a correct result within a specified time called deadline for example the anti-lock breaks on a car are a simple example of a real time computing system, the real- time constraint in this system is the time in which the breaks must be released to prevent the wheel from locking. Whatever its category (hard, soft and firm), a real-time embedded system is said to achieve energy neutral operation if its execution requirements can be supported forever despite energy limitations [8-10]. Now a day’s most of the real-time embedded system executing on the platform that are mobile and carry their own power source in the form of battery and do not depend on power outlet on wall. Most of the time this mobile device remains beyond the scope of recharging their battery due to mobility or unavailability of recharging point, for example mobile video phone applications require light weight device movable across the globe [11, 13]. Most of time these light weighted devices remains beyond the scope of recharges the battery. This may be due to unavailability of recharging point or time required for recharging due to mobility. Thus, for smooth functioning of this light weight mobile device one has to facilitate it either with large capacity battery or powerful power management technique to enlarge the battery back-up time. However in some applications, replacing/recharging a battery is either costly or impractical, wireless sensor net- work is one of such application, where the sensor nodes are deployed in a wide wild area for environment surveillance. Hence, ideally such a system should be designed to operate perpetually with the battery being the only energy source [1, 3]. With the advancement in battery technique harvesting feature is incorporated in battery [4-5]. Harvesting technique is the ideal substitution where recharging/replacing a battery is either costly or impractical. Energy harvesting (also known as energy scavenging) is the process of generating electrical energy from renewable energy sources available in environment. There exist a variety of different energy sources such as solar energy, kinetic energy thermal energy etc. Solar energy is certainly one of the most promising energy source and most of the typical applications have access to solar energy. Therefore, the energy generated by solar panels suffices to execute most common applications equipped with photovoltaic cells, due to which perpetual operation becomes possible without frequent recharging and replacement of the batteries [4]. 1.1 Characteristics of energy harvesting real- time systems 1. Ability to operate by lowest standby current to maximize storage of energy. 2. Consume lowest possible energy when active. 3. Ability to turn on and off instantaneously. 4. Analog capability for sensor interfacing and measurement. 5. Ability to operate with low voltage range. 6. Lowest leakage current to maximize harvested energy. 1.2 Limitations of energy harvesting real-time systems 1. Renewable energy sources available in environment are unstable. 2. Intensity of energy from renewable energy sources varies with time, for example in case of solar energy at day time the intensity of light is very high but during night it will be zero. 3. Limited size of energy storage or battery. Apart from all these constraints, to maximize the possible numbers of tasks to be scheduled, an efficient power management technique is required. Present study focuses on scheduling periodic tasks with deadline, on a uniprocessor platform and variable speed system which is powered by renewable energy storage with limited capacity such as battery or a capacitor. The content of present study is summarized in different sections. Section-2 introduces the related research works. Section-3 describes the motivational example. The energy harvesting system model and some assumptions are described in Section-4. Section-5 explains the proposed methodology with example.