Research On Mobile Communication Network Coverage Analysis System Supporting High- Concurrency

With the rapid development of mobile internet technology,mobile communication network is currently co-operated by 2G,3G,4G carriers,and 5G technology is in the process of preparing for commercial development.How to manage various network resources to cater to the users' high-speed mobile internet service needs and ensure the performance level of data services becomes the core objectives of operators and business platforms.Mobile communication network coverage is the basis of the entire network,which determines the planning,deployment,optimization and maintenance.This topic is based on the Guangzhou Science and Technology Plan project that is for the study and industrialization of 5G network test cloud platform technology,and on the Spark Big Data Platform that has built a mobile network coverage analysis system supporting high-userconcurrency.In this paper,a mobile network coverage analysis system supporting high user concurrency is designed through MR data,DT data collection and analysis of user data in the process of mobile communication.Moreover,all of its features were tested and load has been verified.Firstly,the research background and significance of the network coverage analysis system based on big data platform are introduced.The requirements of deploying Spark big data optimization platform in the current circumstances are discussed.After that,according to the design idea of coverage analysis system,the main technical background is presented,from the system performance test,the working principle and advantages of Hadoop,the working principle and method of Spark,commonly-used network optimization methods and focus,DT&CQT and other dimensionsNext,the design thoughts and implementation steps of the coverage analysis system are explained in detail,including the requirement analysis,architecture design,and module implementation ideas.In addition,it presents the design brain of thought,technical points and implementation methods of the concurrent test module,data cleaning module,coverage analysis module,client side,and automatic.It focus on the LOF density-based distance detection algorithm,DBSCAN clustering algorithm,the classification analysis,query ideas and implementation methods of Spark SQL.Finally,it's said that the software and hardware deployment of the overlay system,tests of each module,regional coverage analysis imitating Zhaotong MR data and automatic road test imitating Guangzhou DT data,whose results are shown on the map interface and able to be exported in excel.In addition,feature detection and performance verification of the coverage analysis system was implemented.The experimental results show that compared with the last network optimization system,this coverage analysis system is more in line with the current big data communication environment since it can process TB-level data,and the processing time reaches the expectation.Compared with other coverage analysis systems on the market,this paper has high compatibility,high adaptability,and high versatility etc.It can reduce the working time and work intensity of network users,improve network efficiency,and it is more in line with the future 5G communication scenarios.