The Study Of Vertical Handoff Based On Multiple Attribute Decision Making Or Fuzzy Logic Control

In the next generation wireless communication system, communication network will be characterized by a variety of wireless access technologies, which provides more access network technologies, different level of Qo S and wireless features to achieve anyone can at anytime, anyplace and related objects to any form of communication with anyone. In order to guarantee the terminal service quality and continuity of the business, need by handoff technology to achieve seamless switching between mobile terminals in the network. Commonly used handoff technologies are mainly included three handoff schemes based on the received signal strength, multiple attribute decision making theory and artificial intelligence. The first scheme decision attributes of a single, consideration is not comprehensive, due to multipath fading is easy to produce ping-pang effect. The second scheme can according to business needs to choose the appropriate decision attribute, flexibility is higher, but considering the decision attribute will have deviation or omissions, and distribution of attribute weights too subjective. Although the third scheme can accurately realize the handoff, the algorithm is too complex to widely use in the general equipment.In this thesis, we mainly aim at the existing problems of handoff algorithms. Three different types of vertical handoff algorithms are proposed. The performances of handoff under different network environment are studied, simulated and verified.(1) A kind of multiple attributes decision making based on Qo S handoff algorithm is proposed, which highlights the importance of service Qo S. We consider the Qo S attributes include bandwidth, delay, jitter and packet loss rate. At the same time, we consider the price attribute in order to satisfy the end user satisfaction. The algorithm mainly combines AHP with TOPSIS, the conversational class, streaming class, interactional class and the background class, for three kinds of candidate networks(UMTS, Wi MAX, and WLAN) to study the vertical handoff.In the Matlab simulation platform, the result shows that the algor ithm is more suitable for more streaming class. Compared with vertical handoff algorithm based on AHP, using the proposed algorithm can reduce the network blocking probability from 16.61% to 8.63% under streaming class.(2) On the basis of the scheme(1), an improved vertical handoff algorithm of multiple attribute decision making, increase considering multiple attributes, such as the received signal strength, terminal movement speed, network load and safety, etc. In the study of attribute weights, we put forward the hierarchical structure. O n the second floor by subjective and objective weights to select attributes total weight value. Using the algorithm can solve the problem that weight allocation is too subjective. In the study of network rank, the proposed algorithm combines with TOPSIS to solve the problem of the abnormal rank.In the Matlab simulation platform, when terminal speed in 1, 5, 10 and 20 m/s respectively, the results showed that compared with the traditional switching algorithm, this algorithm can be adaptive terminal movement speed, when the mobile terminal speed is relatively low, which can choose better access network according to business requirements; When the mobile terminal speed is relatively high, due to stay in some of the network time is too short in small range of coverage would choose a larger network. In session class business, for example, in the case of low speed mobile terminal(1, 5 m/s), terminal switching times for 8 times; And in the case of high-speed mobile terminal(10, 20 m/s), terminal switching times down to 2 times.(3) In this thesis, a vertical handoff algorithm based on fuzzy logic control is used to solve the problems of fuzzy decision- making. For example, speed may have different definition in different network, so the originally precise numerical through fuzzy normalized speed fuzzy control module, make network between direct comparison.Simulation using the Matlab fuzzy logic toolbox in the design of parallel structure of fuzzy logic control structure, which aims to reduce the fuzzy rules and the complexity of the algorithm. Results show that when mobile terminal at a low speed, which can choose the access network according to the type of business. When mobile terminal at a high speed, the algorithm will choose WWAN network have a wider range of communication.