基于自适应分段广延指数模型的IPTV用户点播行为

详细信息    查看全文 | 推荐本文 |

英文篇名：IPTV user's on-demand behavior based on adaptive segmental stretched exponential model 作者：陈步华 ; 陈戈 ; 梁洁 英文作者：CHEN Buhua;CHEN Ge;LIANG Jie;Guangzhou Research Institute of China Telecom Co., Ltd.; 关键词：IPTV ; 点播系统 ; ASSE模型 ; 用户行为分析 英文关键词：IPTV;;on-demand system;;ASSE model;;analysis of user's behavior 中文刊名：DXKX 英文刊名：Telecommunications Science 机构：中国电信股份有限公司广州研究院; 出版日期：2018-05-20 出版单位：电信科学 年：2018 期：v.34 基金：国家高技术研究发展计划(“863”计划)基金资助项目(No.2015AA015803)~~ 语种：中文; 页：DXKX201805018 页数：7 CN：05 ISSN：11-2103/TN 分类号：154-160

摘要

为优化IPTV CDN系统配置,并改善视频调度与存储策略,使得用户获得更优质的视频访问体验,提出了一种基于自适应分段广延指数(adaptive segmental stretched exponential,ASSE)模型的IPTV用户点播行为建模方法。同时,采用中国电信股份有限公司(以下简称中国电信)IPTV实际用户点播行为数据对提出的ASSE模型进行视频访问概率建模仿真。实验表明,与传统建模方法相比,提出的ASSE模型不仅可以达到简化数学模型和减少计算的目的,而且可以更好地提高曲线拟合的精度,与实际数据具有良好的一致性。
        The IPTV user's on-demand behavior model based on adaptive segmental stretched exponential(ASSE) was proposed to optimize the configuration of IPTV CDN system, to improve the video scheduling and storage strategy, and to provide users with better experience for video access. At the same time, the probability of the video access by using the actual IPTV user's on-demand data of China Telecom was simulated. Experiments show that compared with the traditional modeling methods, the ASSE model proposed not only achieves the purpose of simplifying the mathematical model and reducing the computation, but also improves the accuracy of the curve fitting. It has good consistency with the actual data.

引文

[1]张旺俊.Web缓存替换策略与预取技术的研究[D].合肥:中国科学技术大学,2011.ZHANG W J.Research on Web cache replacement strategy and prefetching technology[D].Hefei:University of Science and Technology of China,2011.
    [2]夏琰.基于实际用户行为分析的缓存研究[D].合肥:中国科学技术大学,2011.XIA Y.Research on caching based on actual user behavior analysis[D].Hefei:University of Science and Technology of China,2011.
    [3]韦乐平.三网融合与IPTV的发展和挑战[J].电信科学,2006,22(7):1-5.WEI L P.Triple-play and the development and challenges of IPTV[J].Telecommunications Science,2006,22(7):1-5.
    [4]崔华杰.大型直播与点播IPTV系统的用户行为分析[D].广州:中山大学,2013.CUI H J.Dissecting user behaviors for a simultaneous live and VOD IPTV system[D].Guangzhou:Sun Yat-sen University,2013.
    [5]GUO L,TAN E,CHEN S,et al.The stretched exponential distribution of internet media access patterns[C]//TwentySeventh ACM Symposium on Principles of Distributed Computing,PODC 2008,August 18-21,Toronto,Canada.New York:ACM Press,2008:283-294.
    [6]LAHERRèRE J,SORNETTE D.Stretched exponential distributions in nature and economy:“fat tails”with characteristic scales[J].The European Physical Journal B-Condensed Matter and Complex Systems,1998,2(4):525-539.
    [7]赵晨阳,翟少丹.高尔顿与相关理论的产生[J].西北大学学报:自然科学版,2008,38(4):680-684.ZHAO C Y,ZHUO S D.Galton and the invention of correlation[J].Journal of Northwestern University:Natural Science Edition,2008,38(4):680-684.
    [8]李祖亮.基于数学建模的经济变量线性回归统计预测研究[J].中国校外教育,2011(2):142-142.LI Z L.Research on statistical prediction of linear regression of economic variables based on mathematical modeling[J].Education for Chinese After-School,2011(2):142-142.
    [9]贾丹华,王润润,王鹏.电信套餐资费预演中客户量的预测方法研究[J].电信科学,2011,27(8):25-32.JIA D H,WANG R R,WANG P.Research on prediction method of the amount of customers in Telecom services tariff preview[J].Telecommunications Science,2011,27(8):25-32.
    [10]吕游,刘吉臻,赵文杰,等.基于分段曲线拟合的稳态检测方法[J].仪器仪表学报,2012,33(1):194-200.LV Y,LIU J Z,ZHAO W J,et al.Steady-state detecting method based on piecewise curve fitting[J].Chinese Journal of Scientific Instrument,2012,33(1):194-200.
    [11]方艾,张玉忠,徐雄,等.基于Map R的IPTV用户收视行为分析的方案与实践[J].电信科学,2017,33(2):138-143.FANG A,ZHANG Y Z,XU X,et al.Scheme and practice of IPTV user viewing behavior analysis based on Map R[J].Telecommunications Science,2017,33(2):138-143.
    [12]王重,刘黎明.拟合优度检验统计量的设定方法[J].统计与决策,2010(5):154-156.WANG C,LIU L M.Setting method of statistics for goodness of fit[J].Statistics and Decision,2010(5):154-156.