Application Of Clustering-based Data Mining Technology In Future Network Base Station Deployment Strategy

Under the future network scenario represented by 5G,the deployment of high-density low-power nodes and macro base stations collaboratively is a trend toward base station deployment.Different from the LTE network,the characteristics of high-density deployment makes the site selection of base station even more important.How to use limited resources and lower cost to deploy base stations where they are needed most and how to maximize the benefits of deployment is a topic of concern to network providers.With the rapid development of data mining,machine learning and big data technologies in recent years,the related technologies have become feasible in site-based planning for priority deployment of base stations,especially in future network scenarios featuring high density and low power nodes Program.In this context,this paper describes how to use unsupervised clustering technology to mine user daily traffic data to find areas with high network traffic demand and areas with poor user experience,so as to provide a supplementary basis for site selection of base station priority deployment.At the same time,this paper deeply studies the key technologies involved in data mining.The research scope of this paper includes the data analysis,data preprocessing,clustering algorithm,result interpretation and visualization involved in data mining.This paper elaborates the key technologies and algorithm innovations of each step comprehensively and carefully,and explores some key problems of clustering analysis,including outlier detection,K value selection and clustering quality evaluation,and adaptive K-Means algorithm Innovation.