Research On Cognitive Network State Evaluation

With the rapid development of network technology, the computer networks has infiltrated into every corner of people's lives, and the network quality of service has get more and more attention. Cognitive network is the future development direction of intelligent network. It has the cognitive process and learning function, self-aware, self-optimizing, and self-configuring and other characteristics, it can improve the network quality of service. Cognitive network state evaluation is the basis for cognitive network to cognize, to learn, to plan, to decide and to re-configure. It is also the basis to analyze the existing strategies of cognitive network and design new strategies.This paper first summarizes the concept of cognitive network and key technology, then presents the concept of network state evaluation, describes the methods, metrics framework and evaluation process of network state evaluation. Thirdly, a detailed description of cognitive network state evaluation metrics framework was given, together with the analysis of basic metric selection standard of state evaluation, and the methods and calculation formula of metrics standardization; after the analysis of link state evaluation methods and network integrated state evaluation methods, the evaluation metrics calculation formulas were given. According to the state distribution situation of the whole network, this thesis puts forward network state balanced index. And then presents the index calculation formula. Finally, a experimentation was made to verify the usability and effectiveness of this cognitive network state evaluation metrics framework by NS2 simulation in VoIP network.