Adaptive intelligence job online scheduling within dynamic grid environment based on gridsim

Abstract

This paper concentrates on the design and implement of the grid system for study of adaptive job scheduling algorithm based on GridSim. The common problems of job scheduling in grid system like heterogeneous of jobs, resources and dynamic an arrival time of new jobs significantly changes, can be deal with this solution. The idea behind the adaptive job scheduling algorithm is the hybrid algorithms that consist of Ant Colony Optimization (ACO) and Tabu algorithms. Additionally, the provided common information from Grid Information Service (GIS) and an arrival new job are calculated by Fuzzy C-Means (FCM) algorithm in order· to evaluate the current status of resources and groups of arrival jobs. Moreover, both dynamic and static information are handled by the solution. In static case, the resource information such as a number of CPUs of a machine, CPU speed, a number machine in the grid system is significantly known in advance while dynamic information like the arrival jobs that are submitted to the system any time during simulation. In the results, this paper shows the comparison results between the adaptive job scheduling algorithms and the traditional algorithms.