Font Size: a A A

A Grid Resource Allocation Based On Bidding Mechanism

Posted on:2012-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:W H WangFull Text:PDF
GTID:2218330341450640Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
Resource providers and users have their different goals and policy in grid environment, which makes grid resource allocation a difficulty. How to reasonably allocate resources and optimize system's overall efficiency becomes an important research topic. Most of the traditional resource allocation methods pay close attention to how to improve utilization and allocative efficiency of resources, but rarely consider fairness of the resource allocation, lacking reasonable and targeted allocation algorithm to further improve efficiency and fairness of grid resource allocation.Centering on grid resource allocation, the paper introduces grid related technologies and research status at home and abroad, then analyze advantages of bringing the economics in grid resource management. On the account of the theory above and characteristics of grid resources, with market economic model, a resource allocation model based on bidding mechanism is proposed. In this model, users get resources by adjusting bidding, which makes their own utility maximization. The model firstly designs user's utility function to consider utility optimization with resource competition relations between users and then gets optimal bidding combination of all users. After that, the model allocates resources reasonably according to optimal bidding of users. Thereby, it optimizes the system's overall utility. Meanwhile, a corresponding algorithm is put forward. At last, the algorithm is simulated with the popular simulator GridSim.The result indicates that the resource allocation policy is feasible. Comparing with the other two algorithms, the algorithm improves resource allocation efficiency and fairness in a way.
Keywords/Search Tags:Grid, resource allocation, Utility function, Nash Equilibrium, GridSim
PDF Full Text Request
Related items