A Novel Grid Trust Model Based on Fuzzy Theory

A grid system is a dynamic and distributed environment which contains mass entities and resources, and is involved in large-scale data exchanges and resource sharing. Thus in grid system, information security is a prime concern. Trust mechanism is a main part of grid security. Taking the fuzziness of behavior trust into consideration, this paper introduces fuzzy theory to grid trust and proposes a novel trust model in order to better manage trust relationships between grid entities, presenting the calculation of direct and recommended trust based on fuzzy comprehensive evaluation. Especially, the paper adds the definition mechanism of trust class based on fuzzy cluster analysis to the trust model, achieving dynamic and better visualized definition of trust class. Case study shows that the proposed model can efficiently evaluate entity trust and give clear, accurate judgment upon entities' trust class, which contributes to the establishment of access control policies.