Study On The Policy Of The Integrated Access System Of Communication Grid

Grid is an information technology environment, which is internet-based built and of flexible distribution. Grid is an organic combination of three technologies which are high-performance computers, information sources and internet to support resource sharing and collaborative problem in dynamic network environment. Communication Grid is put forward on the basic concept of grid, which achieves information sharing and management by using the controllability of grid technology. Its main purpose is to achieve the goals of global control and optimize network resources by solving the problems of user admission control, quality of service (QoS), routing security, as well as the allocation of resources.To begin with this thesis, basic knowledge of grid is introduced. On the basis of grid, the concept of Communication Grid and architecture of the Integrated Access System of Communication Grid are presented. Moreover, the functions of the system’s main parts and communication modes of the system are explained respectively. Then, admission control policies of the Integrated Access System of Communication Grid are designed. A communication can be made by anybody by any means in anytime and anywhere through the admission control policies. Furthermore, various of policies can be set up by the policy management interface to ensure the quality of service and reliability requirements of different users and businesses. Besides, a generation method of dynamic routing policies is focused on in this thesis, which is an effective way to implement routing reliability, flexibility and scalability. Routing policies can be generated according to different user, business and network status information.The validity of admission control policies to the new connection is proved, together with the flexibility and controllability of policy mechanism. Besides, different routing policies can be generated under different policy generating conditions, and the effectiveness of routing results in different network state can be improved by using the newly-generated policies.