Taboo-Particle Swarm Optimization Algorithm Based On Trusted Degree For Grid Task Scheduling

Task scheduling is a fundamental issue in achieving high performance of grid computing. Scheduling strategies directly effect the performance of grid. The features of distribution, heterogeneousness, dynamic and self-ruling of grid challenge the traditional scheduling algorithm. Task scheduling algorithms is one of the most popular topics for the Grid technologies research.This paper firstly introduces the conceptual background, category and system structure of Grid. Then discusses the goal, principle and basic strategies of task scheduling in Grid computing. It analyzes the present common grid task scheduling algorithm, and emphatically analyzes some insufficiencies of the PSO algorithm from and the flaw from the two aspects of algorithm' performance and security, using the taboo search algorithm to makes up the shortcoming of PSO algorithm, that is extremely to be easy to fall into partially the most superior in the later period, and introduces the concept of confidence level, proposes one kind of new improvement PSO algorithm, the algorithm first stochastic produces a crowd of granules, then iterates unceasingly the position and the speed of the granule, obtains the task scheduling plan which may be accept, and chooses a group whose confidence level is the highest as superior solution, again obtains the optimal solution by compares the superior solution using the taboo search algorithm. Use the theory way to prove the algorithm accuracy, then did the contrast experiment by the PSO algorithm and the genetic algorithm, use grid dispatch simulation tool GridSim to carry on the massive simulation experiment. The simulation experiment result had indicated that the TTPSO algorithm is more suiting for the solution of big scale grid task scheduling problem compared with the genetic algorithm.At last, the dissertation have a summarization of this research and indicated the direction of its future research.