Data Mining Application In Mobile Network Optimization

Data Mining is a typical Information-oriented intellective application technology. It is not only can analyze and disposal a great deal data but also can discover the relations among the historical data, thereby gets the valuable information. The process of Mobile Network Optimization is to optimize and make good use of the existing network resources. The purpose of Mobile Network Optimization is to ensure and improve the network quality, to improve the enterprise competitiveness and consumer satisfaction. The Mobile Network Optimization is the powerful backup of enterprise development. This article is to research how to use Data Mining technology applied to the Mobile Network Optimization, discuss how to more efficiently optimize the Mobile Network.The Mobile Network Optimization includes the Wireless Network Optimization and the Exchange Network Optimization technology. In the Wireless Network Optimization, Poor Performance Cell optimization is very important, this article puts forward Cluster Analysis Data Mining method to mine and optimize the Poor Performance Cell. The method includes three parts, Firstly we will mine the Poor Performance Cell using Sample Cluster Analysis method, and then we can know the most poor performace cell, At past time we can not realize; Next secondly we analyze the correlativity between the Poor Performance Cell and the performance indicators using variable Cluster Analysis method, mining the more correlative indicators and set down optimization method;Lastly we do network optimization work in time and accurately, and have a definite object in view. The method handles the Poor Performance Cell by way of handling the more correlative indicators, and prevents the Poor Performance Cell appearance by way of controlling the correlative indicators, finally we improve the Wireless Network quality. In the Exchange Network Optimization, the Exchange System Call Connectivity Rate Optimization is the same important, the indicator and Trunk running is correlative. This article analyzes and predicts Trunk Traffic to analyze Trunk running, and then it puts forward regression statistic data mining method to predict the Trunk Traffic. The time(season change mostly) and subscribers in VLR effect the Trunk Traffic, Subscribers in VLR decide the Trunk Traffic development, this article Creates two regression models, One is about Trunk Traffic and Subscribers in VLR, the other is Trunk Traffic and Time,we use the two models to predict the Trunk Traffic, the Trunk resources configuration and utilization are the most optimized and Exchange System Connectivity Rate and Exchange Network quality are improved.These projects are apllied to the Mobile Network Optimization, they breakdown passive statistic pastime, this optimization method adopt scientific calculate and scientific predict, not only it can improve network quality ,but also help product maintenance management department and market management department establish decision-making, and it benefits our company better.