Research On Vertical Handover Decision Algorithm For Heterogeneous Networks

The next generation mobile communication networks will be heterogeneous wireless networks based on convergence and coexistence of various wireless access technologies. A single network is not able to meet the needs ofend users, because of many disadvantages of the existing networks, such as the limited coverage, bandwidth, and not good enough quality of service(QoS) guarantee. In order to ensure that end users can choose an appropriate network in the complex network environment to get a better service with a reasonable price, it is necessary to achieve seamless handover between networks with vertical handover technology. The key to solve this problem is to provide a reliable and efficient vertical handover decision algorithm on which this thesis studies.Firstly, the process of vertical handover is described with emphasis on the existing vertical handover decision algorithms, and the problems that network selection and handover perform timing are not well considered are pointed out.Secondly, an adaptive vertical handover decision algorithm based on fuzzy logic(AVHO-FL) is proposed to deal with these problems. The AVHO-FL classifies context information used in handover decisions into network, terminal, user and service. Next, this thesis analyzes how to determine the handover perform timing, and introduces the concept of handover trend. Then, this thesis uses fuzzy logic to solve the parameter fuzzinessproblem, and proposes a new network selection algorithm,the algorithm gets users’ network like value(NLV) through network status, terminal power consumption, users’ preference and service type, and uses network that has the maximum NLV as the target network to satisfy the demand of network, terminal, user and service. Meanwhile, this thesis presents architecture for vertical handover process to support the proposed handover decision algorithm.Finally, this thesis conducts a simulation and analysis on the proposed vertical handover decision algorithm, and the results show that the algorithm has good performance and can adapt to the requirements of network, terminal, user and service.