Research On Seamless Vertical Handoff Mechanism For Heterogeneous Wireless IP Networks

In the future network development, the wireless network is a kind of seamless integration of heterogeneous network that supports all IP, based on a variety of wireless network access technology. They are not identical in their coverage, access capability, there is no single network in terms of bandwidth, delay, access costs that meet all of the user’s requirements. In order to realize the effective integration of the heterogeneous network, to ensure that make full use of network resources, and move seamlessly between networks. Efficient and independent access mobility management technology and high performance of vertical handoff scheme is the key. Commonly used handoff techniques are mainly based on the received signal strength scheme, the scheme based on artificial intelligence and the scheme based on the theory of multiple attribute decision making. Decision making based on the received signal strength is too single on metrics, and multipath fading is easy to produce the ping-pong effect; multiple attribute decision making choose the appropriate decision attribute according to the business needs, which is highly flexibility, but the decision attribute will have deviation and attribute weights allocation is too subjective; although artificial intelligence can realize the switch accurately, but still hard to use widely due to the complexity of the algorithm.In this thesis, several different vertical handoff algorithms are proposed to analysis performance under the heterogeneous network environment. The fuzzy petri net of context aware Handover decision making algorithm(FPN-CA) is proposed. Firstly, The context information used in handoff decision is classified and summarized, according to distinguish of the mobile devices, network status, type of business and user preference. Put forward the concept of policy vector(PV) and user acceptance value(UAV). Secondly, thesis improves the level of AHP, in order to solve the problem of relatively subjective of user preference, using double level of AHP to obtain the weight of the metrics. In the light of each context metric, thesis presume its user’s preference weights, using combination of AHP and fuzzy Petri net reasoning method, along with user preference for network access point to form the metrics of satisfaction, as the basis for the ranking of candidate network importance. Thirdly, thesis presents a system model of strategy construction, indicate every function module of the system. For the problem that strategy constructing system model can not update context timely when changes occurs. The strategy of automatic construction and update process is presented. Finally, thesis uses MATLAB as simulation platform, analyzes the performance of different types of business data in vertical handoff and compares the proposed algorithm with other algorithms in user acceptance value and network selection to prove the effectiveness of the presented scheme.