Algorithm for deadline based task scheduling in heterogeneous grid environment

Abstract

Grid technology is the structure of technology that provides highly efficient performance in the grid environment. Design of an efficient and reliable task scheduling algorithm is one of the challenging issues in grid computing. A novel improvised scheduling algorithm (IDSA) with deadline limitation for efficient job execution is proposed in this paper. This algorithm is compared with renowned task scheduling algorithms such as Prioritized Based Deadline Scheduling Algorithm (PDSA) and Earliest Deadline First (EDF) in terms of Average Tardiness ARt (the total delay time between the deadline and end time of the task by the total number of tasks) and Number of non-tardy jobs (total number of tasks finishes before their deadline). The proposed algorithm achieves 27.28 % and 30.0 % less ATi than the EDF and PDSA respectively at 4000 number of tasks. The computational results by proposed IDSA for Non-delayed tasks are 2.17 % and 1.70 % higher than the EDF and PDSA respectively at 4000 number of tasks. This enhances the performance of proposed IDSA compare to existing scheduling algorithms and shows IDSA is more suitable scheduling algorithm for grid computing.