基于众包数据的OTT业务行为建模分析  

Modeling Analysis of OTT Business Behavior Based on Crowdsourcing Data

作　　者：李克 刘静怡 宋晓勤[2] 徐小龙 韩涛[3] 

机构地区：[1]北京联合大学信息学院,北京100101 [2]南京航空航天大学电子与信息工程学院,南京210016 [3]华为技术有限公司,南京210012

出　　处：《计算机工程》2017年第10期6-12,共7页Computer Engineering

基　　金：国家自然科学基金(61301103);华为创新研究计划项目(YB2014050020);北京联合大学人才强校行动计划项目(Rk100201510)

摘　　要：针对传统以话音和数据业务为主的业务模型在刻画新型移动互联网业务上的局限性,提出一种OTT业务行为建模方法。基于以众包方式采集的海量真实用户业务行为数据,利用OTT业务行为建模方法对3类典型OTT业务的行为特征进行分析与对比,并对部分业务关键参数的分布特征进行建模。通过分析OTT业务行为的分布特征,可全面揭示出不同移动互联网业务类型之间的行为差异,对于无线网络产品设计、网络规模预测、网络规划和建设具有一定的参考价值。In view of the fact that traditional business models which focus on speech and data services are less effective in describing the business behavior properties of the new mobile Internet business,this paper proposes a new Over-The- Top （OTT） business behavior modeling method. Based on the massive real user behavior data collected by crowdsourcing,the properties of three typical OTT businesses are analyzed and compared with the proposed modeling method, and the distribution characteristics of some key parameters are modeled. Analyzing the distribution characteristics of OTT business behavior can fully reveals the behavior differences between different types of mobile Internet services, which has certain reference value for the design of wireless network products, network scale prediction, and network planning and construction.

关 键 词：业务模型 移动互联网 OTT业务 业务行为 众包 

分 类 号：TP302.7[自动化与计算机技术—计算机系统结构]