Research On Agent-Based Access Control In Distributed System

Internet technology is growing quickly now, independent information organizations have a trend to get together, and more and more enterprises, organizations, governments and peoples need to exchange their information throw these network.. But mistake operations by authenticated users become a big secure threat. So we need to protect network resource through access control. Access control is used to constrain the users' operations on resources. User's permissions decide what they can do. In order to protect those resources could be used under certain rules and reduce secure threat from mistake operations.The aim of this paper is to solve the problem of secure access in distributed system. I adopt RBAC (Role-based Access Control) and agent technology, design and realize access control model in distributed system. Main content is shown below:1. First of all, this paper introduces some basic conceptions of access control; compare some typical access control models. At last I decide to use RBAC as my access control model.2. Second, I analyze Agent technology and it's application. Then discuss the integration of Agent and access Control. Introduce FIPA standard and JADE platform3. Considering the characteristics in distributed system, I use agent and RBAC together to fulfill the need of mutual authentication and guarantee security, integrality and undeniable ness.4. Realizing this model. I use RBAC as inner domain access control policy, and multi-agent system with attribute certificate for authentication and authorization, in order to solve the problem of single sign on in one session.