Grid Resource Selecting Based On Historical Information And Multi-dimensional QoS

It is very difficult to manage the resources in the grid because of the dynamics and the diversities of the resources, then it is better to study the resource discovery and resource selection respectively. The resource selection can choose the appropriate resources for the application from the candidate resources produced by the stage of the resource discovery, and organize it reasonably. In addition, the resource selection is the core of the task scheduling, so it is a hot spot in the grid study currently.With the deepening of the research on the grid, the number of the resources with similar function is increasing, and more and more attention has been paid to QoS (Quality of Service) of the resources. The QoS is defined as a group of nonfunctional attributes, and every element in the group indicates certain quality of the resource. The historical information feedback can reflect the quality of the service, and it is not enough to consider one or two QoS of the resources because of the diversity requests of users. So, the dissertation has done a study on the resource selection based on the historical information feedback and Multi-dimensional QoS constraints.Firstly, this dissertation introduces the related concepts of the grid. Secondly, the current research and the correlative algorithms are expounded. Finally, this dissertation proposes an arithmetic of resource selection based on the historical information feedback (GRSA-HIF), the arithmetic of resource selection based Multi-dimensional QoS (GRSA-MQoS) is also proposed based on GRSA-HIF . Simulation results reveal that the arithmetic of GRSA-MQoS can improve the user's satisfaction and the system's performance respectively.