Intelligent Access Selection in Cognitive Networks: A Fuzzy Neural Network Approach

AbstractTraditional single attribute access selection scheme can’t meet the requirements of new service and userexperience in cognitive networks. In this paper, we present a novel multi-attributes intelligent accessselection scheme based on fuzzy neural network. By means of considering network side, terminal side anduser preference as access elements, we get the optimal access network by using fuzzy neural network.Parameter learning makes the scheme adapt network environment variation dynamically. Simulationresults show the advantages comparing to the existing schemes. Keywords : Cognitive Networks; Access Selection; Fuzzy Neural Network; Particle Swarm Optimization 1 Introduction Cognitive network (CN) is a type of intelligent network that has a cognitive process can perceivecurrent network conditions, and then plan, decide and act on those conditions, then learn fromthese adaptations and use them to make future decisions, all while taking into account end-to-endgoals [1]. Because of its self-management attributions, CN is considered as a key approach toachieve the goals of intelligence thereupon then attract more attention, and become a hotspotthat academia concerned rapidly [2].As one of the key technologies, network access selection plays an important role in CN studies.In CNs, it’s difficult for terminals to select the optimal network rapidly and accurately becauseof networks’ heterogeneous. From recent researches, we can conclude that there are mainly twomajor aspects:1) access selection based on single attribute; 2) access selection based on multi-attributes. Single attribute access selection manner can improve access performance for terminalside and network side at one area. E.g. load balance manner uses load metric parametersto quantify the load degree, and selects access network with lower load [3]. User preferencemanner selects the optimal network according to user preference model, it can improve access