Research On QoS Method Of Cognitive Network Based On Intelligent Theory

With the rapid progress of network access technology and the increasing number of service types, traditional network architecture is faced with more and more challenges. Traditional network is not aware of its whole network information, is not able to adapt to the network dynamics quickly, not to mention reasoning and learning by itself, thus it can not guarantee network quality of service (QoS) effectively. Compared to traditional network, Cognitive Network can sensor the whole network information, make plans and decisions, and take actions accordingly, it is able to reason and learn, and becomes the developing direction of the future network.This thesis aims at guaranteeing QoS over Cognitive Network. At the beginning, background of Cognitive Network and related technologies, such as recent proposed Cognitive Network architectures, DiffServ QoS architecture and service perception technologies are discussed in detail. Based on the Service Type Credit Evaluation Mechanism and Artificial Intelligence, a novel Cognitive Network model, Intelligent Credit Evaluation Cognitive Network (ICE CN) is constructed. Later, a Cognitive Network QoS Guaranteeing Method based on Decision Network is proposed to realize the QoS Guaranteeing Ring in the Triple-Ring structure of ICE CN. ICE CN is compatible and extensible. The simulation results show that ICE CN and its related method are capable of guaranteeing the QoS over network.