Research And Application On Task Scheduling Algorithm Based On Grid

Task scheduling problem is the key of completing large scale calculating assignment in the grid environment,which is necessary foundation of gridding application.High-efficiency gridding assignment scheduling strategy and algorithm can make use of the resources of network system sufficiently and enhance the whole function of gridding application.Because the characteristics of resources of network has plentiful,distributing,heterogeneity and so on,the load and usability of processor in grid environment changes dynamically,and Access Mode and using cost are different, which make the Task scheduling and resources management become more complex.In the environment of grid calculating,how to coordinate and assign network resources,which can make the grid calculating tend to optimization,is the foundation problem of that Task scheduling algorithm has to solve.This issue is on the foundation of analysis classic grid Task scheduling algorithm, from these the Requirement Details of target component which is best span,quality of service,load balance,economic principle and so on,analyze the scheduling strategy, characteristic,run time,Algorithm Limitation and so on of the variety classic Task scheduling algorithm in detail.Using the advantage of ant colony optimization in solving optimization grouping problem designs a grid Task scheduling model and scheduling strategy base on the ant colony optimization and algorithm about them. Adopting the GridSim Toolkit simulation platform tests this algorithm in order to validate the feasibility and validity of the design proposal,and analyzing the experimental result in detail from the point of view about run time,load balance and so on.The experimental result show that the issue adopting the technical of Improved algorithm is feasible and application in the he grid environment.The technology and method in this thesis have been applied preliminarily in the Web-Based Instruction system.