Design And Simulated Implementation Of Intelligent Handoff Decision Mechanisms With ABC Supported In NGI

With the rapid development of Internet changing people’s life and work, more and more variety of new needs have arisen. In order to respond to the change and solve the lack of IP address, Next Generation Internet (NGI) came into being to be able to provide users with high-quality communication services for guarantying QoS and meeting the requirements of Always Best Connected (ABC). In this thesis, intelligent handoff decision mechanisms with ABC supported in NGI are designed in this background.For the purpose of better supporting ABC, user’s QoS requirements, user’s preference on network supplier and coding type, user’s handoff history record, user’s maximum outlay and the profit of suppliers are all considered in the handoff decision mechanisms designed in this thesis. In order to ensure that the QoS parameter’s weight setting of the different types of applications is reasonable, the triangular fuzzy numbers AHP is used. Game theory and Nash equilibrium of micro-economics are introduced to consider the benefits of both the network suppliers and end-users to guarantee the fairness of handoff decision, so that the result of handoff decision achieves all-win. Intelligent handoff decision mechanisms with ABC supported designed in this thesis are based on five intelligent optimization algorithms, namely, free search algorithm based on coarse-grained and fine-grained crossover, quantum-inspired immune clonal algorithm, hybrid IWO/PSO algorithm, improved artificial fish swarm algorithm based on chaotic search and feedback strategy and heuristic kalman algorithm to find the optimal handoff decision for end-users.The above mechanisms are implemented and the performances are analyzed based on NS2environment. The simulation results show that the handoff decision mechanisms designed in this thesis can not only satisfy the user’s QoS requirements, user’s preference and the expectation of outlay, but also consider the benefits of network suppliers. Therefore, the mechanisms can solve the handoff decision problem with ABC supported in NGI effectively.