Comparing the Performance of Real-time Scheduling Algorithms in Terms of their Power Consumption when Working on a Multi-core Processor

Abstract

The exponentially increasing progress of information and communication technology has led to the need to meet the requirements of this technology, of which real-time systems constitute the basic core, in order to ensure a high quality of service in line with the nature of applications spread today, because the main factor in the success of any real-time system is to ensure that tasks are executed before Reaching the time constraint associated with each task.

This led to the need for scheduling algorithms in order to arrange the execution of tasks according to certain preferences so that all tasks are completed without exceeding the time constraint. Researchers in this field have proposed many different algorithms in terms of the principle of attribution of preferences, the nature of the operating platform, and the type of tasks that deal with them.

Most of the research in recent years tends to improve energy consumption in various fields and research applications, and for the real-time field, this means in particular improving the performance of scheduling algorithms so that they provide the desired function with less energy consumption.

In this research, a group of different scheduling algorithms was studied and evaluated in terms of their energy consumption, and energy was measured using a computer running UBUNTU operating system because it contains a set of functions that allow the user to measure energy consumption levels when performing a certain set of tasks, and a scheduling simulation has been carried out Real-time tasks based on SIMSO, an open source simulator that allows simulation of a large number of scheduling algorithms in real time, and is compatible with the UBUNTU operating system environment.