Task Scheduling in Energy Harvesting Real-time Embedded Systems

Harvesting energy from the environment is very desirable for many emerging applications that use embedded devices. Energy harvesting also known as energy scavenging enables us to guarantee quasi-perpetual system operation for wireless sensors, medical implants, etc. without requiring human intervention which is normally necessary for recharging batteries in classical battery-operated systems. Nevertheless, energy harvesting calls for solving numerous technological problems in relation with chemistry when batteries are used for temporary energy storage for example, power management of the embedded computing system that consumes the energy, etc. And this latter problem becomes more complex when the embedded system has real-time constraints i.e. deadlines attached to computations. This paper surveys the main issues involved in designing energy harvesting embedded systems that present strict timing requirements. *Corresponding author: Maryline Chetto, L’UNAM University–IRCCyN (Institut de Recherche en Communications et Cybernétique de Nantes), UMR CNRS 6597, Nantes, France, E-mail: [email protected] Received September 13, 2012; Accepted September 14, 2012; Published September 17, 2012 Citation: Chetto M (2012) Task Scheduling in Energy Harvesting Real-time Embedded Systems. J Inform Tech Softw Eng 2:e106. doi:10.4172/21657866.1000e106 Copyright: © 2012 Chetto M. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Introduction to Energy Harvesting Energy harvesting is a technique that generates electricity from renewable energy sources [1]. Although it has been used for a long time (windmill, solar panel, etc.), the novation lies in the design of smart embedded systems with efficient energy harvesting capabilities while satisfying reliability, dimensioning and timing requirements that characterize modern embedded computers. Any energy harvesting system is composed of three parts which respectively concern conversion from one form of energy in electricity, storage of the energy once it has been harvested from the source and finally consumption of the energy by the embedded computing system. Many forms of energy can be harvested from the environment and consequently they use different types of generators. They include mechanical, thermal, photovoltaic, electromagnetic, biological, and chemical energy. Solar energy and mechanical energy are certainly the most prevalent ones. For example, the mechanical energy can be drained from ambient vibration. Consuming the energy only at the time of its production is not always possible. Consequently, an energy storage unit permits the embedded system to continue operation even when there is no energy to harvest. Generally, rechargeable batteries are used for long term storage. In addition, supercapacitors or ultracapacitors are commonly used for storing transiently the energy from several seconds to minutes without involving aging and rate-capacity problems. Power Management and Task Scheduling Issues Energy harvesting embedded systems have received attention from the research community from about one decade. The main issue in designing an energy harvesting embedded system -besides efficiently extracting and storing the energyis to dynamically adapt the activity of the embedded system to the available energy. In other terms, we have to implement a power management policy that is permanently aware of the operating conditions mainly given by the level of energy currently stored in the storage unit (battery or supercapacitor) and the future incoming energy. The conventional objective of the power management policy in a battery-powered system such as a notebook computer or a cellular phone is to maximize its lifetime under a total energy constraint. There, the focus is on the reduction of the power consumption of the device with Dynamic Power Management (DPM) or Dynamic Voltage and Frequency Scaling (DVFS) [2]. DPM and DVFS techniques reveal to be effective in reducing the energy dissipation but they do not prevent any device to exhaust the battery. In energy harvesting systems, the challenging issue is to operate in an energy neutral mode, consuming only as much energy as harvested. Consequently, this leads to the possibility of indefinitely long lifetime if limitation due to hardware longevity is forgotten. Nevertheless, the objective of energy neutrality can be attained only if certain operating conditions are satisfied, in particular related to the average power generated by the harvesting device and the capacity of the energy storage unit. How to perform power management and how to schedule real-time tasks energyefficiently are consequently the main issues of research in the domain of energy harvesting real-time embedded systems. Real-time embedded systems are characterized by the need for running multiple tasks (programs) on a processing unit such as a micro-controller. Scheduling these tasks on that processor so that real-time constraints are met is a difficult problem [3]. The scheduling problem consists of deciding the order, the start time and the running time of tasks with known characteristics such as deadline, periodicity, duration and energy consumption. We generally classify scheduling policies for real-time systems as static ones when tasks execute in a fixed order determined offline or dynamic ones when the order of execution is decided online. A dynamic scheduling policy is very often based on priority where priority intuitively represents a measure of the urgency of each task. Earliest Deadline First (dynamic priority depending on urgency) and Rate Monotonic (fixed priority depending on period) can support sophisticated task set characteristics such as deadlines, precedence constraints, shared resources, jitter, etc., [4]. Despite the significant body of results in real-time scheduling, many real world problems are not easily supported, including energy harvesting.