Heterogeneous Laxity-based Improved Task Scheduling for DAG-based Heterogeneous Computing

Conference: McGraw-Hill International Conference on Signal, Image Processing Communication and Automation
Author(s): Rashmi Basavaraddi, Arshiya Sultana Year: 2017
Grenze ID: 02.MH-ICSIPCA.2017.1.45 Page: 288-294

Abstract

In heterogeneous computing taskmapping is necessary and has drawn major attention. Task scheduling algorithms\nat present are less efficient. This generally consists of two phases where first is pritorization and second is task assigning. In\nthis paper a Heterogeneous laxity based improved task scheduling algorithm is proposed. This method of scheduling is\nrepresented by directed acyclic graph (DAG). This combines the method Heterogeneous Laxity Based Scheduling (HLBS)\nand Heterogeneous Scheduling with Improved Task Priority (HSIP) to improve the task scheduling efficiency. Here in this\npaper we come across with m-processers and n-tasks and we are concentrating on ideal time slot of any m-processers to\nschedule any n-tasks with shorter laxity and execution time to complete the task. To address the performance issue we are\nconsidering three parameters like makespan, Scheduling Length Ratio (SLR) and failure ratio.

<< BACK

MH-ICSIPCA - 2017