新疆移动精细化业务规划研究

发布时间：2018-08-25 13:00

【摘要】：新疆移动通信业正处于2G增速放缓、3G高速发展、LTE商用起步的关键阶段。一方面用户规模和业务量快速增长；另一方面用户对移动通信业务种类的需求的多样化也在不断加强。新疆移动业务需求规划方案设计的合理性能确保预判未来移动通信市场的业务需求量，对新疆移动公司的业务发展、网络规划建设等做出科学的、有数据支撑的指导和建议。目前，国内业务规划主要以定性方法进行人为的估计分析方法来推断移动业务发展情况，该方法缺少数据支撑致使结果说服力不强，无法满足当前精细化业务规划的要求。本文对新疆移动业务的现状进行分析，面对不同的业务采用不同的预测方法，首次采用自上而下与自下而上的方法进行预测，对运营收入、用户数、计费时长和数据流量的预测结果进行对比验证。在预测过程中，针对输入数据维数太多，并且数据之间相关性不强等原因，对数据进行了预处理。数据预处理主要工作包括相关性分析、K-L信息量法、格兰杰因果检验、主成分分析和聚类分析等定量分析方法。本文采用多元线性回归模型等预测方法，建立了预测模型预测某地区未来三年的业务数据。最终，本文以一个市级地区的数据进行了实验验证，通过对运营收入、用户数、计费时长和数据流量的预测，得到该地区未来三年运营收入、用户数、计费时长和数据流量的发展需求。本文所完成的某区公司的规划试点工作成果，所得到的运营收入、用户数、计费时长和数据流量的预测结果在向新疆移动公司的汇报展示中，，获得了新疆移动公司的充分认可，为下一步全疆的业务规划和网络规划做出必要的基础准备和经验累积。
[Abstract]:The Xinjiang mobile communication industry is in the critical stage of 2G growth slowing and 3G high-speed development and LTE commercial start-up. On the one hand, the user size and business volume are growing rapidly; on the other hand, the diversification of users' demand for mobile communication services is also increasing. The reasonable performance of Xinjiang Mobile Service requirement Planning Scheme ensures that the future mobile communication market needs to be forecasted, and provides scientific and data-supported guidance and suggestions to Xinjiang Mobile Company's business development, network planning and construction. At present, the domestic service planning mainly uses the qualitative method to carry on the artificial estimation analysis method to infer the mobile service development situation, this method lacks the data support causes the result is not convincing, cannot satisfy the present fine service planning request. In this paper, the current situation of Xinjiang Mobile Services is analyzed. In the face of different services, different forecasting methods are used for the first time, the top-down and bottom-up methods are used to forecast the operating income and the number of users. The forecasting results of charging time and data flow are compared and verified. In the process of prediction, the data is preprocessed because of too many dimensions of input data, and the correlation between data is not strong. The main work of data preprocessing includes correlation analysis K-L information method Granger causality test principal component analysis cluster analysis and other quantitative analysis methods. In this paper, the multivariate linear regression model is used to predict the business data of a certain area in the next three years. Finally, this paper has carried on the experimental verification with the data of a city level area, through the forecast to the operating income, the user number, the billing time and the data flow, obtains the operation income, the user number in the next three years in this area, Billing time and data flow development requirements. The results of the planning and pilot work of a certain district company, the forecast results of operating income, number of users, billing time and data flow achieved in this paper have been fully recognized by Xinjiang Mobile Company in its report to Xinjiang Mobile Company. Make the necessary basic preparation and experience accumulation for the next business planning and network planning of Xinjiang.
【学位授予单位】：新疆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F626

【参考文献】