Journal Article

Application of Heuristic Algorithm in Accomplishing Computer Instructions

Abstract

Focusing on the Infrastructure as a Service (IaaS), scheduling seems to be the most primary problem of cloud computing as tasks (cloudlets) scheduling have to be carry out simultaneously and efficiently in order to meet the cloud consumers’ expectations. To realize this goal, tasks have to be executed as fast as possible in order to reduce the total completion time (makespan) and ensuring efficient resource (virtual machine) utilization. Many algorithms had implemented to remedy this problem. Algorithms such as Max-min is implemented which finds the tasks with maximum execution time and allocates to the resource with it minimum completion time. The limitation of Max-min scheduling algorithm is of high make span and low resource utilization when the number of tasks with maximum execution time is more than the number of tasks with minimum execution time. To address this problem, a new algorithm called Max-Range is proposing which allocates tasks based on range of completion time. A tool kit known as cloudsim (cloud simulator) was used to implement and compare the performances of the proposed algorithm with existing algorithms. The results of the experiment indicate that the proposed algorithm is able to produce good quality solutions by reducing the make span and reduce the cost of execution as compared to the standard Max-min scheduling algorithm

Download pdf