| Nowadays, the communication network is more and more challenged by the efficient Quality of Service (QoS) management of increasing complexity. The concept of cognitive network (CN) appears as a candidate that can address this issue. It is a new type of data network that makes use of gathering technologies from several research areas (machine learning, knowledge representation and network management etc.) to advance the performance of the network for getting end-to-end QoS goals. The CN learns from the history of the network, and uses this knowledge to improve the decisions in the future.As a critical part of CN, the knowledge plane (KP) spans over layers of the whole OSI model, therefore having global knowledge as well as local's of the network and its elements. It is a pervasive system within the network that builds and maintains high level models of what the network is supposed to do, in order to provide services and advice to other elements of the network.In this thesis, with the requirements of the future network, we summarize the recent researches of CN, and then design an extendable architecture of KP, which is capable of making appropriate decisions with its knowledge to help network elements adjust themselves for guarantying end-to-end QoS goals.This thesis firstly analize the situation of current network, the demands in the future, and an extendable QoS guaranteed architecture of CN, to ensure the importance and features of KP in CN. Secondly, it designs a process of network knowledge cognition in multiple levels, and an extendable QoS guaranteed architecture of KP. Then we focuses on network traffic to implement the architecture. Thirdly, we prepare a scenario for the implementation, and then the result of the scenario test validates that the architecture meets the design requirements.In the end, we summarize the whole thesis. Current shortcomings of the architecture design and future work are given. |
PDF Full Text Request