| Grid Computing System is an integrated cooperative computing environment that is composed of geographically distributed heterogeneous resource. Grid Computing breaks through the location of geography and the means of traditional cooperation. We should share various resources more widely and solve more complicated problems.Job scheduling of Grid is a NP-hard problem. The inherent parallelism and scalability of ant colony algorithm make the algorithm very suitable for job scheduling in the Computational Grid. In this paper, we apply the ant colony algorithm to solve the job scheduling problem in Grid and bring forward a job scheduling implement through the movement of the job-agents.The algorithm updates local pheromone when the jobs are assigned to the grid resources, and updates global pheromone after these jobs are executed. By simulation experiment, we tested and selected parameters of ant colony algorithm, and got fair results. Actually, the algorithm makes the job dispatcher and load balancing effectively.The main work of this paper includes the following aspects:(1) First of all,the paper introduces the concept of the Grid,the characteristic of the Grid,the significance of the Grid and the system framework of the Grid.(2) Sceondly, the paper introduces the concept ,the goal ,the characteristic of Job Scheduling and some Scheduling algorithms in the Computational Grid.(3) By comparing the similarities and differences,the paper discribles the characteristic and the bug of the Basic Ant Colony System, as well as some improved algorithms,such as Ant System with elitist strategy, Rank-Based Version of Ant System, Ant Colony System,Max-Min Ant System and so on. (4) Finally, the paper introduces the base ideal of Job Scheduling based on Ant Colony Algorithm and we design and implement the Job Scheduling Algorithm in Computational Grid. By the simulations,we fix on the best combination of arithmetic parameters,and analyse the validity of the algorithm in job scheduling. |
PDF Full Text Request