摘要
随着Internet的快速发展和个人计算机性能的提高,人们不再仅仅满足传统的文字网页浏览、邮件收发、在线聊天等应用。由于流媒体特有的娱乐性和交互性,它越来越受到人们的青睐,并且成为目前因特网应用的最快增长点。
本文通过对CDN和P2P优缺点及目前CDN与P2P主流融合方式的研究,提出了基于CDN与P2P的混合网络架构HCPMSA (Hybrid CDN-P2P Streaming Media Service Architecture),另外对影响系统稳定性和工作效率的关键技术进行了研究。
本文的具体研究内容和研究工作如下:
(1)研究了CDN与P2P技术各自的优缺点,及现在的研究的两种技术的主要融合方式。设计了一种即可靠又易管理同时又具有可扩展性的基于CDN和P2P混合网络架构的体系结构。
(2)研究了影响系统稳定性的主要因素,通过对各个因素的分析本文建立了系统稳定性的数学模型,通过系统稳定值把系统分为三个等级的稳定状态。
(3)研究了目前流行的P2P节点选择算法,建立了可靠度的数学模型;提出了一种基于可靠度的P2P资源节点选择算法,此算法综合考虑了服务节点的可靠度、负载均衡及贡献度的问题。
(4)研究了流媒体编码扩展技术,系统把流媒体文件分为不同的等级。把用户请求分为不同的优先权,提出了一种基于系统稳定性的分级预防与恢复策略。该策略中的分级指的是媒体文件分级、用户优先权的分级及系统稳定性的分级。
本文的实验基于Cygewin+NS2环境,测试了系统的吞吐量、系统中节点的抖动率及请求成功率,结果显示这几方面都有明显的提高。说明本文设计的方案能够在保证一定程度服务质量的的同时提高系统的稳定性。
本文的创新点如下:
(1)设计一个利用了CDN的可靠性、可管理性和P2P技术的可扩展性的CDN与P2P技术的网络架构。
(2)根据系统的负载状况和网络状况提出了系统稳定值的概念,设计一个基于系统稳定值的分级预防与恢复调度策略,设计了一个基于可靠度的P2P节点选择算法此算法综合考虑了节点负载均衡、节点贡献度、节点可靠度的问题。
With the rapid development of Internet and PC performance improvement,people no longer just satisfied with traditional text web browsing, email transceiver and online chatting application. Because of the peculiar entertainment and interactivity of streaming media, it has more and more get the favor of people, and become the fastest growth in the current Internet application.
In this paper, based on the research of advantages of CDN and P2P and current mainstream fusion way, we proposed the mixed network architecture with CDN and P2P HCPMSA (Hybrid CDN-P2P Streaming Media Service Architecture), in addition, the key technology which effect stability of the system and the efficiency is discussed.
The special research content and work are as follows:
(1) We have made comprehensive research on the advantages, disadvantages of CDN and P2P, and the main fusion way of the two technologies now. A reliable, easily managed and extensible mixed network architecture with CDN and P2P.
(2) We study the main factors influencing system stability. The mathematical model of analysis of the factors is established in this paper, which sets the system to a 3 hierarchy through the stable value.
(3) The popular P2P node selection algorithm is investigated in this article, and a mathematical model of reliability is established; a P2P resource node selection algorithm, based on the reliability, comprehensively considering the reliability, load balance and contribution of service node, is put forward.
(4) We research streaming media coding expand technologies. In the system, streaming media files are divided into different grades, and the user requests are divided into different priorities. A strategy based on the stability of the system classification prevention and rehabilitation is proposed. The grading in the system refers to the media file classification, users of priority and the stability of the system hierarchical classification.
The experiment environment is based on Cygewin+NS2. It mainly tests the system throughput, jitter rate and request rate of the peers. The experiment result shows great improvement in these parameters. This proves that the design of this system can guarantee the QoS as well as improve the stability of the system.
The innovation points of this paper are as follows:
(1) It gives a network architecture of CDN and P2P which combines the reliability and manageability of CDN and the extendibility of P2P.
(2) According to the load situation and network condition, it puts the concept of system stability value and designs a prevention and scheduling strategy based on it.It designs a peer choice algorithm based on reliability. The algorithm considers peer’s load balance, contribution value and reliability at the same time.
引文
[1] Lazar I,Terrill W.Exploring content delivery networking[M].IT Professional,2001,3 (4):47-49.
[2] Molina B,Ruiz V.et al..A closer look at a content delivery network implementation Electrotechnical Conference[C].2004 MELECON 2004,Proceedings of the 12th IEEE Mediterranean,2004,685~688.
[3]闫旭. CDN互联网新兴网络技术的研究[J].电视工程,2007,(02):30-34.
[4]王松.中国CDN的发展与ChinaCache的成长之路[J].中国传媒科技.2005,6-17.
[5] Xuguang Lan, Nanning Zheng, Jianru Xue et al.. A Peer-to-Peer Architecture for Efficient Live Scalable Media Streaming on Internet[C]. Proceedings of the 15th international conference on Multimedia,2007:783-786.
[6]唐瑞春,刘斌,徐仕儒.分层P2P系统中基于簇缓存的节点选择策略[J].中国海洋大学学报.2010,40(1):104-108.
[7] Lei Guo,Songqing Chen,and Xiaodong Zhang. Design and Evaluation of a Scalable and Reliable P2P Assisted Proxy for On-Demand Streaming Media Delivery[C]. IEEE transactions on knowledge and data engineering,2006,18(5):669-682.
[8] Dongyan Xu,Sunil Surech Kulkami et al.. Analysis of a CDN -P2P hybrid arachitecture for cost-effective streaming media distribution[J].Multimedia System,2006,11(4):383-399.
[9]唐瑞春,魏青磊,刘斌.一种基于P2P协作的代理缓存流媒体调度算法[J].电子与信息学报.2009,31(11):2757-2761.(EI: 20095112567102)
[10]杨传栋,余镇危,王行刚等.结合CDN与P2P技术的混合流媒体系统研究[J1.计算机应用,2005,25(9):2204-2207.
[11] X Zhang, J Liu, B Li et al.. Coolstreaming/DONet:A data-driven overlay network for peer-to-peer live media streaming [C]. Proceedings of IEEE INFOCOM, Miami, USA, 2005: 2012-2111.
[12] Vikash Agarwal , Reza Rejaie. Adaptive multi-source streaming in heterogeneous peer-to-peer networks [C]. Proceeding of Multimedia Computing and NetWorking,San Jose,California,2005: 151-160.
[13]杨戈,樊秀梅.基于CDN的流媒体动态调度算法[J].通信学报,2009,30(02):42-46.
[14]郑常熠,佘宇东,王新等. CDN与P2P混合网络架构下基于用户行为的VoD分发策略[C]. 2007年通信理论与技术新发展, 2007: 2027-2032.
[15]郑婕,张松,齐洁等. P2P流媒体节点选择机制的研究与仿真[J].计算机工程与设计2007,28(22):5396一5399.
[16]冯健,房鼎益,潘亚峰等.一种混合式P2P流媒体点播系统中的节点选择算法[J].西北大学学报(自然科学版),2008,38(2):199-216.
[17] Xu D, Hefeeda M, Hambrusch S, Bhargava B. On peer-to-peer media streaming[C]. Proceedings of IEEE ICDCS. 2002,35-43
[18]杨明川. CDN中的负载均衡技术[J].信息网络,2003,(21):27-28.
[19]卓力,沈兰荪,朱青等.视频流关键技术的研究进展[J].电子学报,2002,030(008)
[20]张洋,张楠,尹宝才等.多描述编码研究现状[J].计算机学报,2007,30(9):1612-1622。
[21]黄晓涛,张纬,朱华等.基于P2P流媒体的分层多描述编码[J].计算机科学,2009,36(12):65-69.
[22]昌庆江,邹静娴,纪志成.多描述分层编码技术的研究[J].电视技术,2005,(6):16-18.
[23] Wang H,Ortega A. Robust video communication by combining scalability and multiple description coding techniques[C]. In proc. Of SPIE Image and Video Communications and Processing,San Jose,CA,2003,1(6):111-124.
[24] Tang Rui-chun,Wei Qing-lei,and Liu Bin.A Proxy-Caching Prefix Assignment Algorithm Based on P2P Cooperation in the Media Streaming System [C].2009 International Conference on Multimedia Information Networking and Security,Wuhan,China,2009.658-662.(EI:20101012761333)
[25]管昌生,邓磊,崔华.基于P2P技术的流媒体服务模型研究[D].武汉:武汉理工大学学报, 2006. 26-32
[26] Rainer Lienhart ,Matthew Holliman.Improving Media Services on P2P Networks[J],2002,6(1):73-77
[27] Reibman A R,Jafarkhani H,Wang Yao et al.. Multiple descrip-tion coding for video using motion compensated prediction[C],Image Processing,1999.ICIP 99.Proceedings.1999 International Conference. 1999,(3):837-841
[28] Apostolopoulos J G. Reliable video communication over Lossy packet networks using multiple state encoding and path diversity[M]. Visual Communications and Image Process. 2001,(6):392-409
[29]唐瑞春,纪红英,巩存群.基于簇的移动多媒体客户请求调度策略[J].中国海洋大学学报.2011,(1/2):169-174
[30] Le Y C,Yang L T,Jiang Q et al.. Measurement and analysis of BitTorrent[J]. Computer Research and Development,2008,45(9):1589-1600.
[31] Llie D,Popescu A. Statistical models for Gnutella signaling traffic[J]. Computer Networks,2007,51(17):4816-4835.
[32] Stoica I,Morris R,Liben N et al.. Chord: A scalable peer-to-peer lookup protocol for Internet applications[J]. IEEE/ACM Transactions on Networking,2003,11(1):17-32.
[33] Rowstron A and Druschel P. Pastry: Scalable, decentralized object location, and routing for large-scale peer-to-peer systems[J]. Lecture Notes in Computer Science,2001,22(18):329-350.
[34]崔莉. CDN技术及其在媒体点播系统中的应用研究[D].南京:东南大学,2005.6-12
[35]汪成军.基于CDN的Mediaswitch流媒体系统关键技术研究[D].长沙理工大学硕士学位论文,2007.30-35
[36]余晓俊.CDN-P2P混合模式流媒体平台的研究和实现[D].上海:复旦大学硕士学位论文,2008.8-10.
[37]徐卫东,王康.适用于内容分发网络的动态负载均衡策略[J].计算机科学,2005,32(1):41-44
[38] Tolga Bektasa,Osman Oguza,Iradj Ouveysi. Designing cost-effective content distribution networks[J]. Computers & Operations Research,2007,(34):2436-2449.
[39]彭湘凯.CDN网络及其应用[J].微计算机信息.2005,21(2):150-152.
[40] Hao Yin,Xuening Liu,Tongyu Zhan et al. Design and deployment of a hybrid CDN-P2P system for live video streaming : experiences with LiveSky [C]. International Multimedia Conference ,Beijing,China ,2009,25-34.
[41] Shi Kang,Hao Yin. LiveSky : Enhancing CDN with P2P [J]. ACM Transactions on Multimedia Computing, Communications, and Applications,2010,6(3):1-19.
[42] Hao Yin,Xuening Liu,Tongyu Zhan et al. A Hybrid CDN-P2P System for Video-on-Demand [C]. Proceedings of the 2010 Second International Conference on Future Networks,2010,309-313.
[43]常可沛.基于P2P和CDN技术的分层流媒体系统及其关键问题研究[D].厦门:厦门大学硕士学位论文,2008. 9-14.
[45] S.GM.Koo,C.S.G.Lee,and K.Kannan. A Gnetie-Algorithm-Based Neighbor-Seleetion Strategy for Hybrid Peer-to-Peer Networks[C]. Proe of the 13th IEEE International Conferenceon Computer Communieations and Networks(ICCCN’04),Chieago,IL,2004,469-477.
[46] Chu YH,Rao SG,Zhang H. A case for end system multicast[C]. Proc Of ACM SIGMETRICS,Santa Clara,2000,1-12.
[47] Paul F,Sugih J,Cheng J·IDMaPs. A global internet host distance estimation service[J]. IEEE/ACM Transaction on Networking,2001,9(5):525-540.
[48] NgTSE , Zhang H. Predicting Internet network distance with ecordinates-based approaches[C]. Twenty-First Annual Joint Conference of the IEEE Computer and Communieation Soeieties,2002,170-179.
[49] John Jannotti,David K Gifford,Kirk L Johnson,et al.. Overeast:Reliable multieasting with an overlay network[C]. Proceeding of the Fourth Symposium on Operating Systems Design and ImPlementation,USA,2000:197-212.
[50]唐瑞春,徐仕儒,魏青磊.代理协助P2P-VoD系统中基于缓存状态的服务节点选择策略[J].中国海洋大学学报.2010,40(4):099-104.
[51] Hefeeda M,Habib A,BotevB et al.. PROMISE:peer-to-Peer media streaming using ColleetCast[C]. Pro Of the1lth ACM Int’l,Conf on Multimedia,New York,2003,45-54.
[52] Zhang XY,Liu JC,LI B et a1.. Coolstreaming/DONet:A data-driven overlay network For live media streaming [C].Proe of the IEEE INFOCOM,Miami,2005,2102-2111.