Research on the Adaptable Reputation-Based Scheduling for Storage Resources in Data Grids

Data grids provide the data-intensive applications with a large virtual storage framework. We use a reputation-based trust aggregation approach for data providers to quantitatively predict and calculate whether or not they are reliable and credible. In the same time, adaptable trust service is increasingly important for applications running on data grids. However, existing systems are unable to satisfy the flexible trust needs of complex data-intensive applications. This paper introduced a novel approach and proposed a quality of reputation-based trust framework for storage resources in data grids (QRTF). QRTF leverages the manager to dynamically control quality of trust for application requests, thereby achieving good tradeoffs between reputation-based trust service and system performance. Experimental results based on a simulated grid with storage resources show that the proposed framework is capable of significantly improving trust service quality and guaranteeing desired response time.