Study On Trust-based Access Control In An Open Environment

The purpose of this article is to realizae resource access control of computer system in the open network environment. We consider the problem than it is difficult to apply resource access control because we can not validate the user’s identify in open network environment. We analysis the relationship between trust and access control, then we put forward a method which using the human subjective trust as factors in access control. We analysis the research progress of access control models both at home and abroad, then we put forward the trust based access control model.First of all, we put forward the subjective trust evaluation model based on fuzzy theory. According to the trust characteristics of subjectivity, fuzziness, we use fuzzy correlation theory and method of mathematics to express trust level. We use fuzzy comprehensive evaluation method to measure trust quantitatively, and we define attribute tree of trust as a factor evaluation sets. We use 7 levels of trust as the evaluation set, and give a new definition of fuzzy operator according to the factor set and evaluation set of synthetic fuzzy relation matrix. We use trapezoidal function as membership function to calculate the degree of membership, and use analytic hierarchy process to calculate the weight of each trust factors.Secondly, We put forward the trust level based access control model. We add trust level as a factor of access control on the base of role-based access control model.Each trust level has a set of roles. We add some rights to a role, and we set up a series of permissions for each role. We describe the whole progress that a user obtain the resource by trust level detailed.Finally, according to the design of the proposed theoretical model we implement a simple resource access control system. We achieve a variety of algorithms proposed in the access control model in the system, and provide function of adding and accessing resources. Each user can set the role and authority of their own resources, the weight distribution of the trust of each factor is also calculated by each user themselves according to their own habits.