Research On Network Selection Algorithms Of Heterogeneous Wireless Networks Based On Fuzzy Logic

With the continuous development of the mobile Internet,users' business has evolved from normal mobile communication to multi-request business.Nowadays,there are two main mobile network techniques: mobile communication system and wireless local area networks system.They both have advantages and disadvantages,one can't replace the other completely.Therefore,current wireless network environment is the convergence of heterogeneous wireless networks,which contains multiple network access points and various wireless networks.To guarantee users an excellent business experience in such environment,it's the hot and difficulty spot of the convergence of heterogeneous wireless networks that how to select and handoff to the best network adapted to users' requests.Handoff decision is the key of vertical handoff,and the core is the network selection algorithm.An outstanding network selection algorithm can both guarantee users best QoS anytime in a dynamic heterogeneous wireless networks environment and least influence on users' business when handoff executing.In this paper,two network selection algorithms of heterogeneous wireless networks based on fuzzy logic are proposed.The first algorithm is Network Selection Algorithm based on Normal Fuzzy Numbers(NSANFN),it replaces traditional triangular fuzzy numbers in Fuzzy AHP method with normal fuzzy numbers,overcoming the inherent defects of triangular fuzzy numbers.It also uses Normal Fuzzy Extent Analysis Method(NFEAM),which combines normal fuzzy numbers and fuzzy extent analysis method,to calculate weights of network attributes.Finally rank the networks and select best network with TOPSIS.According to the experiments,the algorithm is more accurate,and the weights calculated from it are more suitable for human.The second algorithm is Network Selection Algorithm based on Fuzzy Inference in Dense Networks(NSAFIDN).Fuzzifier should be firstly executed with normal membership function to transform input network attribute parameters to fuzzy sets.Then execute fuzzy inference to get the comprehensive evaluate fuzzy set that network belongs to according to fuzzy rule base.Finally defuzzifier will be executed to transform fuzzy set to precise value,with which ordered sequence can get.The algorithm will choose one of the networks that have biggest precise value.The experiments show that the algorithm can effectively shorten the computing time as well as select the handoff network ranked in the front by TOPSIS,achieve the goal of algorithm.