| Summary: | 碩士 === 國立中山大學 === 資訊工程學系研究所 === 107 === A cloud computing platform nowadays has to deal with a large number of user requests at the same time. As such, an efficient workflow scheduling algorithm—which take into account as many factors as it can, such as rent cost, transmission cost, and execution time—will certainly have a strong impact on the performance of the cloud computing platform. This thesis presents such a workflow scheduling algorithm based on a new metaheuristic algorithm, which is referred to as search economic, along with the so-called 2-opt local search mechanism. The results indicate that the proposed algorithm provides an efficient solution to minimize the makespan and the total cost, thus enhancing the overall performance of a cloud computing platform.
|
|---|
