Research On Handoff Decision Algorithm For Improving User Experience Based On Data Mining

With the continuous development and innovation of wireless communication technology,a kind of heterogeneous wireless network with multiple wireless access technologies has been formed.Especially with the gradual deployment of 5G microcells,the heterogeneous wireless networks are becoming more and more dense.Networks with different access technologies bring more choices to people,so how to improve user experience while meeting the basic needs of users in ultra-dense heterogeneous wireless network switching has become a hot issue in academia and industry.Therefore,this thesis proposes corresponding solutions for improving user experience from the perspectives of mining user consumption preferences and predicting network security based on data mining technology.The main work of this thesis is as follows:1.Aiming at the issue of how to improve user of user experience while the essential needs of services are met,a vertical handoff algorithm based on personalized consumption preference in heterogeneous wireless networks has been proposed.First of all,relevant parameters meeting basic needs of services of networks are obtained by defined threshold and the process of parameters in the preprocessing module.Next,personalized consumption preference distribution of users are acquired via data mining module in the server.Last but not least,a customized utility function is constructed by combining acquired parameters and the distribution,and select target network through comparing utility values of networks.The simulation demonstrates that proposed algorithm makes sure that users are able to access network meeting personalized consumption preference on the premise of satisfying the basic needs of services,and reduce the expenditure of time and handover blocking rate and improve total network throughput and user experience.2.Aiming at the problem of how to improve user experience on the basis of enabling users to access a relatively safe network while meeting consumer consumption preferences in an ultra-dense heterogeneous wireless network with fake base stations and fake AP points,a data mining-based predictive network security handoff algorithm is proposed.The main idea is: firstly,data collection and threshold screening are performed to obtain a candidate network that meets the user's basic needs;Secondly,the user's main consumption preferences obtained from the first research point is combined to calculate the corresponding network performance indicators;Thirdly,in order to obtain network security indicator,security prediction to candidate networks is performed using data mining tools;finally,a new customized utility function is constructed based on the user's main consumption preference weight and corresponding network performance indicators and network security indicators,and the best access network among the candidate networks is determined.The simulation results show that the algorithm can enable users to access a relatively safe network while meeting the user's consumption preference,reduce the handover blocking rate,increase the total network throughput,and thereby enhance the user experience.