Reasearch Of Routing Mechanism In Cognitive Network Architecture

With the development of network technology, people's life has changed a lot. There will be different services in the future network, which of course will be compose by different network such as Internet, Ad-hoc network etc. The traditional network structure couldn't meet many new coming demands. Researchers are now developing new network architectures. Cognitive network is one of those directions, it can implement self-* features of network such as self-management, self-configuration, self-optimization, self-healing etc. What's more it has certain cognitive ability, which is able to learn the network status and make network adjustment actively.This thesis firstly introduces the current research situation of network development, and then discusses the architecture of cognitive network, which was divided into Data Plane, Distribute Plane, Discovery Plane, Decision Plane and Knowledge Plane, and each plane is briefly described.Routing is the basic of network connective. While the services in network become diversification and each type has its QoS requirements, this thesis focus on the routing mechanism supports QoS of different services. After a survey of routing techniques the routing architecture in cognitive network is discussed. Then a routing model was proposed under the background of cognitive network.In order to support the multi-class QoS routing, this thesis proposes a routing mechanism based on service-agent which named CAMP (Context Aware Multi-class Routing Protocol). A service-agent support the routing of correspond type of service. The network resources are dynamic shared according to the Local Path Information. As the simulation proved, this mechanism could guarantee the QoS of high priority services while minimize the influence of which done to low priority services. What is more, it can balance the traffic distribution and reduce the chance of congestion, using the network resource effectively.