Research On Access Selection Of Heterogeneous Networks Based On Fuzzy Logic

As users’ expectations for communication quality are getting higher and higher,a single network can not meet the needs of all network users,heterogeneous networks came into being in this environment,a variety of networks overlapping coverage in the signal range,complementing each other in wireless technology,making up for the shortcomings caused by a single network,heterogeneous network convergence has become an inevitable trend in network development.How to select the most suitable network for access in heterogeneous networks has become an important research problem in the development of heterogeneous networks.Based on the comprehensive understanding of the existing wireless heterogeneous network access selection algorithms,this thesis proposes a heterogeneous network access selection algorithm based on fuzzy approximation ideal value method and a heterogeneous network access selection algorithm based on competitive learning particle swarm optimization fuzzy neural network.Aiming at the existing heterogeneous network access selection algorithm,although it can choose the appropriate network according to the network status,but there is still the problem that it cannot be selected according to different user service types,this thesis proposes a heterogeneous network access selection algorithm based on the combination of fuzzy logic and approximate ideal value method.When calculating the weight,the CRITIC method is used to calculate the objective weight of the alternative network,considering the user characteristics of different service types,the subjective weight of the network attribute is calculated by the fuzzy analytic hierarchy method,and in order to reflect the preference of the user type,the coefficient of the subjective weight is higher than the objective weight when the combined weight of the network attribute is obtained by the simple weighting method.Finally,the fuzzy approximation ideal value method is used to calculate the score of each candidate network and sort the network.The simulation results show that the algorithm can not only ensure that different types of users can access the most suitable network,but also reduce the number of switching between networks and the occurrence of ping-pong effect.Aiming at the problem that the existing heterogeneous network access selection algorithms cannot be adaptively adjusted in the face of dynamically changing network information,this thesis proposes a heterogeneous network access selection algorithm based on competitive learning particle swarm optimization fuzzy neural network.When faced with uncertain information in the network environment,the algorithm uses fuzzy neural networks for network selection,which can adjust parameters adaptively.At the same time,in order to improve the accuracy of network evaluation,the competitive learning particle swarm algorithm is used to optimize the initial value of membership degree in the neural network.Considering the influence of user preference on network selection,set the preference value of different users for the alternative network,comprehensively weighted the network score and the alternative network preference value,and use the final weighted value as the basis for network selection.The simulation results show that the algorithm can improve the network throughput,reduce the blocking rate,and reduce the number of network switches while ensuring user satisfaction.