Research On The Technology Of Admission Control Based On Theory Of Complex Networks

Admission control originates from the wired network, and as an effective tool for congestion control and QoS, its mechanism has been extensively studied. As the increasing demand of market and network technology, the numerous sorts of business and dynamic complex network environment raise a higher requirement for the compatibility of admission control mechanism. Complex network is a hot research topic in recent years, which offers a new and powerful way to study the complexity of various networks. How to effectively employ the theory of complex network into admission control technology appears an important research topic, while very rare work has been done worldwide.In this thesis, we tend to employ the theory of complex network into admission control technology, and propose a Complex Network based admission Control (CNAC). CNAC considers the distinction of biases on the same business from different customers, as well as the distinction of biases on different business from the same customer, and construct utility function to transfer different biases into corresponding degrees of QoS. Another contribution of this paper is to propose the concept of degree of activity based on the scale-free network model, which is used to evaluate the mutual attracting and connecting abilities between network nodes. Based on these, we realize the admission control process that corresponds to businesses of various superiorities, and simulate the results on Matlab. The simulation results indicate that, by considering the different bias of various customers and mutual attracting abilities between network nodes, the proposed admission control algorithm improves the utility rate of system band and accepting rate of various businesses, compared with classic counterpart.In this thesis, we also introduce the structure of three-layer cognitive network, which the author has studied in the national 863 projects, and analyze the feasibility and reasonability of applying our proposed admission control algorithm on this cognitive network.