Study On Trust Model In Grid Envioromnet

The grid technology is an up-to-date technology in the area of wide-area network computing, especially in the area of wide-area high-performance computing, which enables coordinated use of geographically distributed resources. The grid technology is thought to be the next wave of improvement of information technology, which is also considered to be the next generation internet.The grid has the following characters: resources in the grid are geographically distributed, resources and clients are connected through wide-area network such as internet; the categories and quantities of resources are tremendous, and resources need to cooperate with each other to some extent; resources are dynamic including the attributes changes, the migration and replication in the grid; resources work in heterogeneous platforms administrated by different policies.However, the characters of grid environment bring many security problems. The security problems in grid environment are being paid more and more attention to. Trust problem in grid environment is a hot topic in currently grid research. It refers to the trust management between consumer and service provider.In order to resolve the security problems due to the special characters of grid environment, in this paper a new trust model is developed, which insists on multiple agents in each domain. The entities are distributed across these agents, with each of these entities being associated with more than one agent. By doing this, the model can not only shortens the response time between entities, but also improve the success rate of transaction. And the model can resolve security problems existing in grid environment more effectively, and can be verified by simulated experiments.Multi-Agent system brings in load-balancing problem.To address this problem, a new and simple strategy which based on redirect payment mechanism is developed, and the effectiveness of the mechanism is proved.The anasys result indicates that the mechanism can resolve the load-balancin problem and enhance network idle resource utilization.