电信运营商基于hadoop平台的大数据系统研究与设计word论文.docxVIP

摘要移动互联网时代，客户消费形态发生巨大变化，移动互联网流量发生爆炸式增长使得电信运营商除了面对传统领域对手的竞争，还需要应对诸如腾讯、百度、新浪等新兴互联网企业的挑战。并且随着电信运营商由“管道提供商”向 “互联网业务提供者”转型，需要电信运营商整合移动互联网行为数据，提供移动互联网下的客户行为分析等应用；整合终端信息数据，提供移动互联网业务环境下，客户与终端关联分析、终端与套餐关联分析等应用；挖掘企业数据价值，提升营销支撑能力。本文针对中国电信现有数据分析系统的存在的问题进行分析和研究，根据集团公司企业信息化部ITSP3.0的要求，对中国电信集团公司系统现状进行总结，并结合目前市场策略和业务特点进行新一代hadoop系统进行分析设计。包括对数据采集、ETL、ETL调度管理、多维分析、领导视窗、自助取数、重点指标监控等业务的业务流程梳理，对数据分析系统总体架构、软件功能的设计进行了详细的阐述，对ETL采用Hadoop技术进改造进行了详细的说明，对系统数据库等关键应用的开发进行了详细的描述，并对系统功能性测试和性能测方案进行了设计，最后介绍了该系统的部署实施和上线运行情况。关键词：大数据HadoopETLvAbstractThemobileInterneterais comingandcustomer consumption patterns changed dramatically.Theexplosivegrowth ofInternet trafficmakes telecom operators facingnot only rivals from thetraditional areas, but also other emergingInternet business challenges such as Tencent,Baidu , Sina and so on.And thetelecom operators who need to transfer from pipe provider to Internet serviceprovider ,havetointegrate mobileInternetbehavior data, provide themobileInternet applications such as customer behavior analysis, integrate terminal information data, provide correlation analysis of customers with theterminal ,correlation analysis of packages with terminals based on mobileInternet business environment ; miningenterprise datavalueand enhancemarketingsupportcapabilities.In this article, it do the analysis andresearch on ChinaTelecomsData analysissystem. existingproblems. Basedon the ChinaTelecom data analysis, system function and performancerequirements areput forward.Overalltheplan forChinas telecom data analysis system and the detailed design of systemfunction, includingthe process ofdatagathering, extract transformingandloading(ETL),ETLschedulingmanagement, multidimensional analysis, leader’s windows, Self-serviceaccess, keyindicators to monitor.Data analysis system overall architecture, function ofthe softwareis been designed in detail.ITdescribed thefunction transfom from SMPto Hadoopdetail.It described theapplication ofthe system database, the development has carried on the d