蜂窝网络的无线资源预测方法与平台实现

发布时间：2018-05-07 13:13

本文选题：蜂窝网络 + 无线资源　；参考：《北京邮电大学》2016年硕士论文

【摘要】：随着移动通信产业的高速发展,蜂窝网络已步入大规模商用4G时代,移动用户数量和业务量指数增长,而移动互联网技术的发展和数据采集技术的逐步完善也促使运营商和设备商可以采集到更多维度、更加详尽的复杂网络数据。随着蜂窝无线网络大数据时代的到来,如何结合大数据技术挖掘蜂窝网络中的海量数据,如何科学有效地利用已有数据进行无线资源管理,从而应对新形势下网络建设与优化的挑战,已成为时下的一个研究热点。LTE网络拥有截然不同的话务模型和业务模型,传统的数据分析平台已经不适用如今的蜂窝无线网络。本论文着重研究了蜂窝网络的无线资源预测方法以及新的数据分析平台实现。论文的主要内容如下:第一、基于国内典型城市的LTE蜂窝网络数据,详细分析了数据的类型、特点和规律。细致调研和总结了无线资源预测相关的典型数据挖掘算法,重点介绍了聚类算法和时间序列预测算法的原理和特点,深入研究了这两类算法在通信领域的具体应用。第二、依照聚类和时间序列预测两类数据挖掘算法,结合蜂窝网络无线资源的特点,提出了一个基于聚类的蜂窝无线资源时序预测模型。模型首先利用聚类算法对基站进行分类,再对典型类型的基站利用时序预测算法进行针对性无线资源的预测,从而找出各类基站的优选预测算法。同时,详细介绍了数据处理,建模和预测的处理流程,通过聚类结果对基站进行针对性的时间序列预测,总体准确率可以提升15%以上。第三、基于4G蜂窝网络数据特点和理论模型建立了一个面向蜂窝网络无线资源数据分析平台,依托传统数据库技术及大数据技术设计数据仓库从而实现了各大典型城市数十亿条级别数据的实时分析和处理。基于对海量蜂窝无线资源数据的详细分析和深入挖掘,平台不仅能够对无线资源基础数据做多维度的可视化展示,还结合相关数据挖掘算法动态地给出预测结果及相关图表,揭示了无线资源数据间的内部规律和关联,为网络建设和优化提供了有力的参照依据。
[Abstract]:With the rapid development of mobile communication industry, cellular network has entered the era of large-scale commercial 4G, and the number of mobile users and business volume has increased exponentially. The development of mobile Internet technology and the gradual improvement of data acquisition technology also promote operators and equipment to collect more dimensions, more detailed complex network data. With the arrival of the era of big data, how to mine the massive data in the cellular network, how to use the existing data scientifically and effectively to manage the wireless resources, In order to meet the challenges of network construction and optimization under the new situation, LTE network has become a research hotspot. LTE network has different traffic model and service model. The traditional data analysis platform is no longer suitable for today's cellular wireless network. This paper focuses on the cellular network wireless resource prediction method and the implementation of a new data analysis platform. The main contents of this paper are as follows: first, based on the LTE cellular network data of typical cities in China, the types, characteristics and laws of the data are analyzed in detail. The typical data mining algorithms related to wireless resource prediction are investigated and summarized in detail. The principles and characteristics of clustering algorithm and time series prediction algorithm are introduced, and the specific applications of these two algorithms in the field of communication are deeply studied. Secondly, according to two kinds of data mining algorithms, clustering and time series prediction, combined with the characteristics of wireless resources in cellular networks, a clustering based time series prediction model for cellular wireless resources is proposed. The model first classifies the base stations by clustering algorithm, and then uses the time series prediction algorithm to predict the targeted wireless resources, and then finds out the optimal selection and prediction algorithm of all kinds of base stations. At the same time, the processing flow of data processing, modeling and prediction is introduced in detail. The overall accuracy can be improved by more than 15% through clustering results to predict the time series of base stations. Thirdly, based on the characteristics of 4G cellular network data and theoretical model, a wireless resource data analysis platform for cellular network is established. Based on the traditional database technology and big data technology, the data warehouse is designed to realize the real-time analysis and processing of billions of data in typical cities. Based on the detailed analysis and deep mining of massive cellular wireless resource data, the platform can not only visualize the basic data of wireless resources in many dimensions, but also dynamically give the prediction results and related charts combined with the related data mining algorithm. The internal rules and correlation of wireless resource data are revealed, which provides a powerful reference for network construction and optimization.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.13;TN929.5

【参考文献】