基于蜂窝网络实测数据的业务流量分析与优化的研究

发布时间：2018-02-10 10:03

本文关键词： 移动互联网资源占用实测数据网络优化　出处：《北京邮电大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着国内3G网络和智能终端的快速发展,移动网络的数据业务流量迅猛增长,不断推出的新形式数据业务,都给运营商的网络运营带来了前所未有的冲击。数据业务的多样性、随机性和突发性的特点,对无线资源的占用不能像传统语音业务一样使用爱尔兰进行衡量和分析,但是如果在进行网络容量规划时不能充分估计数据业务的无线资源占用情况,将会影响用户的使用感受,如何更加科学的规划未来的数据业务发展北重,需要探寻不同数据业务对无线资源的占用规律。本论文分别对WCDMA网络和CDMA20001x EVDO网络中的实测数据进行分析研究,二者侧重点有所不同。针对WCDMA网络,本文主要的研究点,一是选取四种典型的数据业务：QQ、UC、微博和优酷,分别对它们的包长分布规律进行研究,拟合出相应的分布函数曲线；二是结合现有的研究成果和实测数据分析业务对网络资源的占用情况。本文的主要研究重点和创新点是对EVDO网络的研究,主要分三部分：业务分析、资源映射分析和网络优化。业务分析包括了对业务流量的时间分析,分析现有经典流量模型,结合现网数据,分别建立了时间流量模型。资源映射分析主要是针对采集到的现网数据,利用关联映射算法,建立业务资源映射关系函数,建模的结果表明关联算法具有一定的准确性和科学性,建立的映射函数能够表征出不同业务对网络资源的占用规律,对网络优化有一定的指导意义。网络优化则是基于业务资源映射函数提出了三种优化方案：网络用户容量优化、单业务网络资源优化和全网业务网络资源优化,通过这三种网络资源优化方案能够为决策者提供更加清晰明确的依据。概述本文的创新点主要有：对四种典型数据业务的包长分布规律进行分析,拟合分布函数；构建了一种业务流量时间模型；利用现网实测数据进行网络资源占用分析,建立了一种业务资源映射函数；基于映射函数提出了三种网络优化方案。
[Abstract]:With the rapid development of domestic 3G network and intelligent terminal, the data traffic of mobile network is increasing rapidly, and the new data service has brought unprecedented impact to the network operation of the operator. Because of its randomness and paroxysmal characteristics, the occupation of wireless resources cannot be measured and analyzed using Ireland as traditional voice services, but if the wireless resource occupancy of data services cannot be fully estimated in network capacity planning, How to plan the future development of data services more scientifically, it is necessary to explore the rules of different data services to the occupation of wireless resources. In this paper, the measured data in WCDMA network and CDMA20001x EVDO network are analyzed and studied, and the emphasis of the two is different. For WCDMA network, the main research points of this paper are as follows: firstly, four typical data services: QQQUC, Weibo and Youku are selected. The distribution law of packet length is studied respectively and the corresponding distribution function curve is fitted. The second is to analyze the occupation of network resources by combining the existing research results and measured data. The main research focus and innovation of this paper is the research of EVDO network, which is divided into three parts: service analysis, resource mapping analysis and network optimization. Combined with the existing network data, the time flow model is established, and the resource mapping analysis is mainly aimed at the collected existing network data, using the association mapping algorithm to establish the business resource mapping relation function. The results of modeling show that the association algorithm is accurate and scientific, and the mapping function established can represent the rules of the occupation of network resources by different services. Network optimization is based on the service resource mapping function. Three optimization schemes are put forward: network user capacity optimization, single service network resource optimization and whole network service network resource optimization. Through these three network resource optimization schemes, we can provide more clear and clear basis for decision makers. The main innovations of this paper are as follows: analyzing the distribution law of packet length of four typical data services, fitting the distribution function, constructing a traffic flow time model, analyzing the occupation of network resources by using the measured data. A business resource mapping function is established and three network optimization schemes based on the mapping function are proposed.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.53;TP393.06

【参考文献】