Energy Efficient Algorithms for Slack Reclamation
Loading...
Files
Date
2015-12-18T07:21:05Z
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In this project, we present few energy-cognizant algorithms using slack reclamation method. In the present design, the Earliest-Deadline-First (EDF) scheduling policies have been employed on the both CPUs for the dynamic tasks and FIFO scheduling for static tasks. When a new task arrives and the primary processor is active then that task will be scheduled in secondary if that processor is inactive else based on priority context switching will happen. The proposed System shows significant energy saving compared to existing SSPT (Standby Sparing for Periodic Tasks) algorithms.
In Energy-Aware Standby-Sparing Technique for Periodic Real-Time Application, the application tasks are executed on the primary processor using DVS [3]. The spare does not use DVS and is reserved for executing backup tasks, if needed. Upon the successful completion of a primary task, the corresponding backup task is cancelled and excessive energy consumption is avoided. However, if the primary task fails, the backup runs to completion at the maximum frequency. Hence, the solution statically constructs the schedule on the spare CPU to delay the backups as much as possible to minimize the overlap between the primary and backup copies of the same task.
Description
Keywords
Energy Efficient Algorithms for Slack Reclamation, Preethi Radhakrishnan, Praveen, Pradeepti Chowdhury, Asha Kumari