The Research Of Grid Resource Management And Task Scheduling Algorithms Based On Muti-QoS

Grid is the strategic infrastructure of intense computing and information services, which has become the key technique of the next generation of internet. In recent years, the high speed networks develop very fast, which makes it possible to integrate the distributed and heterogeneous computing resources into Computational Grid.It provides an ideal platform to solve large-scale computing problems in scientific and engineering area. Because the resources in Grid are distributed, heterogeneous and dynamic, how to manage the resource and schedule tasks in Grid to meet users' requirements is a challenging problem.With the proliferation of the Grid, quality of service (QoS) has become an important factor that needed to be considered in the procedure of scheduling.In this situation, the hierarchical structure of Grid QoS is analzed and designed first, and then muti-QoS fuzzy attributes are merged into the management of resource and task scheduling algorithms to make the grid tasks execute on the uncertain environment on time.This paper has the following contribution:1. First of all, the concept, objective of grid computing and the architecture of grid QoS are presented; and then the importance of task scheduling model based on QoS is introduced. And on the basis of above, the detail current research situation of grid teconology and tasks scheduing are analyzed and summarized.2. An interval fuzzy multiple QoS measurable approach based on the D-S theory is put forward to deal with the fuzzy and uncertain of user's reqirement, which can merge together all of the QoS submitted by users.The requirements of users will become a priority queue finally. The experimental result shows that it can satisfy the tasks scheduling based on multiple QoS reqirements of users by intervals.3. To discover the resource on tasks scheduling more efficient, the rational management of the grid resource is very important.An improved DBSCAN Clustering algorithm based on Data Mining for partition heterogeneous resource into many similar sets according to their QoS properties is an effective way of manageing the resource.4. Considering the multiple QoS factors of grid environment, the current grid scheduling algorithm is improved and some new scheduling algorithms which take the makespan as the priority objective are presented. The experimental result shows that they all work well in different kinds of heterogeneous envirement. 5. Referring to the social trust relationship model and considering the QoS trust between the grid nodes, an interrelated task scheduling algorithm making the QoS as the priority objective is presented. The algorithm taking the tasks order and communication into consider is more in line with the actual grid. The Simulation shows that the algorithm can efficiently satisfy the QoS requirement with costing a few more time.In the end, the conclusion is made and the problem encountered is analyzed, and the future research is put forward.