Based Grid Resource Trading And Optimization Of The Economic Theory

The heterogeneous, dynamic and autonomous of grid resourced, it has brought great difficulties to the grid resource management and trade. Market economics model is one of the effective means to the grid resource management. There are many of the related work. Three kinds of grid economic models and their characteristics, The General Economic Grid Architecture (GESA), Grid Architecture for Computational Economy (GRACE) and Business Grid model in IBM, were described in this paper.It is still Inadequate in the existing economic grid model, the resource management architecture in the early presented work, it is only resolved the problems such as heterogeneous, when cross multiple administrative domains, but it was lack of a deeply study in meeting needs and improving multi-user resource allocation performance. The attributes of Quality of Service, try one's Best, it can not meet the requirements. The trust of both sides, the application of the price mechanism is difficult to obtain a balance between supply and demand of resources, how to determine the price of resources is the optimal, and so on, there are many issues to be resolved. In early grid QoS scheduling, many methods were based on local optimization, they take into account the QoS information provided by the grid users, but it needs further study to address the overall QoS optimization problem.In order to solve these problems, the author starts from the analysis of the current economic model, proposes the QoS-based grid banking model. It describes the basic principle, structure and execution processes in the grid banking model. Different levels of QoS properties are researched, QoS parameters are classified. In the model, the author puts forward an updated price algorithm, designs cross-layer QoS resource optimization model. Simulation results show the feasibility of the optimization algorithm. It is the conclusion and future work in the last part.The principal innovations in this paper are:1. It puts forward the QoS -based grid banking model, describes its structure, QoS attributes, classification and processes at different levels.2. It proposes an update price based on the algorithm of Deadline and Budget Constrained, through analysis the usage of resource in history, it can get the expected price in the next time .this method is more conducive to safeguarding the interests of resource providers in the grid system, and it will make resources more effective used.3. It proposes cross-layer QoS resource optimization models, optimization algorithm based on the economic principle, the global resource optimization problem decomposed into three sub-problems, and continuous interaction between agents is repeated until all of the participants have reached a common solution.