Research And Simulated Implementation Of Qos Routing Mechanisms Based On Swarm Intelligence With Abc Supported

With the continuous expansion of Internet, new technologies are constantly developed to meet the new network requirements. Therefore, there is a tendency of rapid combination of various wired and wireless networks such as computer, TV, telecommunication, and satellites. The concept of NGI (Next Generation Internet) is introduced at this time. It provides people better network services with higher QoS (Quality of Service). However, some characteristics of NGI, such as heterogeneity and uncertainty, bring troubles for providing reliable QoS. Concerning this problem, many mechanisms and protocols need to be redesigned. In the NGI, when users are routing, an ABC (Always Best Connected) connection must be kept while users change their access points. That is to say, during the whole process of communication, users always enjoy the "best" connective state and the "best" access mode.In this thesis, QoS routing mechanisms based on swarm intelligence with ABC supported are focused, considering QoS of user communication and influence of NGI links' characteristics on QoS routing comprehensively. The inaccuracy of parameters is solved by fuzzy mathematics methods. The fairness problems between users and network providers are handled through related knowledge in microeconomics. By designing strategy sets for users and network providers and calculating the utilities of users and network providers, the strategy combination for users and network providers with Pareto optimality is found under Nash equilibrium. By doing this, the best utilities between users and network providers can be assured and the all-win of the interests of both users and network providers can be achieved. Furthermore, concerning the utility calculation of users and network providers, users' preferences are considered in the links which can be chosen, so that the features of ABC can be fully displayed in communication process. When solving the problem of QoS routing with ABC supported, there are multiple objects needed to be optimized at the same time in the routing mechanisms, which makes the problem into a multi-constrained QoS routing problem. The swarm intelligent algorithms are used to solve it.QoS routing mechanisms based on swarm intelligence with ABC supported designed in this thesis are simulated on NS2 (Network Simulator 2) platform. Some test schemes are proposed for the performance evaluation of the routing mechanisms in this thesis. In these test schemes, the performances of the routing mechanisms are evaluated comprehensively, by analyzing the data of the optimized objects in the routing mechanisms. Compared with some of the existing QoS routing mechanisms, the routing mechanisms introduced in this thesis can provide fairer guarantee. In the heterogeneous, fuzzy and uncertain network environment of NGI, these mechanisms can support QoS in real network communication more effectively...