Ku-Mahamud, Ku Ruhana and Ramli, Razamin and Yusof, Yuhanis and Mohamed Din, Aniza and Mahmuddin, Massudi (2012) Ant colony optimization algorithm for dynamic scheduling of jobs in computational grid. Project Report. Universiti Utara Malaysia, Sintok. (Unpublished)

PDF Restricted to Registered users only Download (1MB) | Request a copy

Abstract

Computational grid is gaining more importance due to the needs for large-scale computing capacity. In computational grid, job scheduling is one of the main factors affecting grid computing performance. Job scheduling problem is classified as an NP-hard problem.Such a problem can be solved only by using approximate algorithms such as heuristic and meta-heuristic algorithms.Among different optimization algorithms for job scheduling, ant colony system algorithm is a popular meta-heuristic algorithm which has the ability to solve different types of NP-hard problems.However, ant colony system algorithm has a deficiency in its heuristic function which affects the algorithm behavior in terms of finding the shortest connection between edges.This research focuses on a new heuristic function where information about recent ants’ discoveries has been considered.The new heuristic function has been integrated into the classical ant colony system algorithm.Furthermore, the enhanced algorithm has been implemented to solve the travelling salesman problem as well as in scheduling of jobs in computational grid.A simulator with dynamic environment feature to mimic real life application has been development to validate the proposed enhanced ant colony system algorithm. Experimental results show that the proposed enhanced algorithm produced better output in term of utilization and makespan in both domains.

Item Type:	Monograph (Project Report)
Additional Information:	Kod S/O: 11986
Subjects:	Q Science > QA Mathematics > QA76 Computer software
Divisions:	College of Arts and Sciences
Depositing User:	Prof. Dr. Ku Ruhana Ku Mahamud
Date Deposited:	09 Dec 2014 09:17
Last Modified:	09 Dec 2014 09:17
URI:	https://repo.uum.edu.my/id/eprint/12789

Actions (login required)

View Item