基于P2P的VoD系统的研究与实现
|
摘要
近年来,随着宽带网络、大规模存储以及高性能工作站等技术的突飞猛进,互联网上传输的信息已不再只是简单的文本和图像,各式各样的多媒体信息服务从经济上和技术上成为可能。VoD (Video on Demand)即视频点播开始盛行起来。
在互联网上提供大规模的VoD服务是一项极具挑战性的工作。传统的VoD是基于C/S (Client/Server)集中式网络模式的,面临网络带宽受限、视频服务器负担过重、支持的用户数目有限以及节目播放效果较差等多方面的问题。服务器的性能成为了限制整个系统服务能力的瓶颈。近几年来P2P文件共享系统的盛行,为我们提供了一个新的思路。P2P技术打破了传统的C／S模式,在网络中的每个节点的地位是对等的。每个节点既充当服务器的角色,为其它节点提供服务,同时也可以享用其它节点提供的服务,很好地解决了C／S模式中的服务器网络带宽瓶颈问题。而BitTorrent协议是一种被广泛应用的P2P文件共享协议,近几年在世界上广泛流行,它的可靠性和高效性都得到了充分的检验。
本文提出并实现了一个基于P2P的VoD系统PPVoD,它将流行的BitTorrent P2P文件传输技术和一个传统的视频服务器结合在一起,通过P2P网络节点之间数据的共享来降低服务器负载压力和提高客户端的数据获取速度。
BitTorrent协议是用来进行文件共享的协议,在很多方面并不适合VoD系统。针对其分发视频数据的低效性,PPVoD系统对BitTorrent协议做了多方面的优化和改进,加入了特殊的视频服务器节点来增强服务质量,提出了随机选择和顺序选择相结合的混合片段选择策略来确保数据获取的高效性及视频播放的流畅度,并拓展了数据缓冲区的管理提出了以最多替换算法为核心的客户端缓存管理机制,同时将BitTorrent中的Tit-for-Tat下载激励机制成功的应用于VoD环境来提高系统效率
本文在NS2平台下,针对视频的播放流畅度、服务器依赖度等主要性能指标对PPVoD进行了仿真实验。具体地试验并讨论了不同的片段选择算法、缓存机制以及用户行为等对系统性能的影响。由实验结果分析得出,基于改进的BitTorrent协议设计的PPVoD是可靠的、高效的,相较传统的VoD系统,BitTorrent协议的加入和算法的改进极大的降低了节点对服务器的依赖程度,用户的QoS满足率也明显提高。本文提出的架构设计以及实现和实验方案为P2P视频点播系统性能优化的研究提供了有价值的参考方案和实验数据。
With the spread of broadband communications and the rapid development of network technology, contents on Internet have gradually changed from simple texts to multimedia data including text、audio and video. VoD(Video-on-Demand) has become one of the greatest potential value-added services on Internet.
Traditionally streaming media VoD applications with poor scalability are based on C/S model. It may easily make server the bottleneck and can not effectively support the large-scalable streaming data distribution. When P2P (peer-to-peer) thinking is introduced to streaming media technology, VoD system use the common nodes' bandwidth、storage and other resources to serve other nodes, so that it reduce loads of server and network greatly and has scalability and decentralized services.
This paper presents and evaluates PPVoD, a scalable Video-on-Demand streaming system that combines the popular Bit-Torrent peer-to-peer file-transfer technology with a simple dedicated streaming server to decrease server load and increase client transfer speed. The system includes a modified version of BitTorrent that supports streaming data delivery and that communicates with a VoD server when the desired data cannot be delivered in real-time by other peers.
However, BitTorrent Protocol is not originally designed for time-sensitive content distribution and indeed is not suitable for time-sensitive data distribution. In this dissertation, we make modifications to BitTorrent Protocol in the following aspects. First, in order to guarantee high QoS to users, media server is added in the system. Second, instead of the original rarest-first piece pick policy, We propose a hybrid schedule algorithm of order-first and beta random with probable selection. At last, Owing to high bandwidth of streaming media, the PPVoD system extends the buffer management and puts forward the buffer algorithm.
In addition, we make an experiment on PPVoD system and design the simulation on NS2 according to the main performance indicators of media distribution。The PPVoD system plays a good performance in media distribution according to the experimental results. The system not only presents little dependent on the media server, but also shows low latency and high media playback fluency. In one word, the dissertation makes a foundation to the optimization for P2P streaming system performance and provides valuable reference to the experimental results and the solutions.
在互联网上提供大规模的VoD服务是一项极具挑战性的工作。传统的VoD是基于C/S (Client/Server)集中式网络模式的,面临网络带宽受限、视频服务器负担过重、支持的用户数目有限以及节目播放效果较差等多方面的问题。服务器的性能成为了限制整个系统服务能力的瓶颈。近几年来P2P文件共享系统的盛行,为我们提供了一个新的思路。P2P技术打破了传统的C／S模式,在网络中的每个节点的地位是对等的。每个节点既充当服务器的角色,为其它节点提供服务,同时也可以享用其它节点提供的服务,很好地解决了C／S模式中的服务器网络带宽瓶颈问题。而BitTorrent协议是一种被广泛应用的P2P文件共享协议,近几年在世界上广泛流行,它的可靠性和高效性都得到了充分的检验。
本文提出并实现了一个基于P2P的VoD系统PPVoD,它将流行的BitTorrent P2P文件传输技术和一个传统的视频服务器结合在一起,通过P2P网络节点之间数据的共享来降低服务器负载压力和提高客户端的数据获取速度。
BitTorrent协议是用来进行文件共享的协议,在很多方面并不适合VoD系统。针对其分发视频数据的低效性,PPVoD系统对BitTorrent协议做了多方面的优化和改进,加入了特殊的视频服务器节点来增强服务质量,提出了随机选择和顺序选择相结合的混合片段选择策略来确保数据获取的高效性及视频播放的流畅度,并拓展了数据缓冲区的管理提出了以最多替换算法为核心的客户端缓存管理机制,同时将BitTorrent中的Tit-for-Tat下载激励机制成功的应用于VoD环境来提高系统效率
本文在NS2平台下,针对视频的播放流畅度、服务器依赖度等主要性能指标对PPVoD进行了仿真实验。具体地试验并讨论了不同的片段选择算法、缓存机制以及用户行为等对系统性能的影响。由实验结果分析得出,基于改进的BitTorrent协议设计的PPVoD是可靠的、高效的,相较传统的VoD系统,BitTorrent协议的加入和算法的改进极大的降低了节点对服务器的依赖程度,用户的QoS满足率也明显提高。本文提出的架构设计以及实现和实验方案为P2P视频点播系统性能优化的研究提供了有价值的参考方案和实验数据。
With the spread of broadband communications and the rapid development of network technology, contents on Internet have gradually changed from simple texts to multimedia data including text、audio and video. VoD(Video-on-Demand) has become one of the greatest potential value-added services on Internet.
Traditionally streaming media VoD applications with poor scalability are based on C/S model. It may easily make server the bottleneck and can not effectively support the large-scalable streaming data distribution. When P2P (peer-to-peer) thinking is introduced to streaming media technology, VoD system use the common nodes' bandwidth、storage and other resources to serve other nodes, so that it reduce loads of server and network greatly and has scalability and decentralized services.
This paper presents and evaluates PPVoD, a scalable Video-on-Demand streaming system that combines the popular Bit-Torrent peer-to-peer file-transfer technology with a simple dedicated streaming server to decrease server load and increase client transfer speed. The system includes a modified version of BitTorrent that supports streaming data delivery and that communicates with a VoD server when the desired data cannot be delivered in real-time by other peers.
However, BitTorrent Protocol is not originally designed for time-sensitive content distribution and indeed is not suitable for time-sensitive data distribution. In this dissertation, we make modifications to BitTorrent Protocol in the following aspects. First, in order to guarantee high QoS to users, media server is added in the system. Second, instead of the original rarest-first piece pick policy, We propose a hybrid schedule algorithm of order-first and beta random with probable selection. At last, Owing to high bandwidth of streaming media, the PPVoD system extends the buffer management and puts forward the buffer algorithm.
In addition, we make an experiment on PPVoD system and design the simulation on NS2 according to the main performance indicators of media distribution。The PPVoD system plays a good performance in media distribution according to the experimental results. The system not only presents little dependent on the media server, but also shows low latency and high media playback fluency. In one word, the dissertation makes a foundation to the optimization for P2P streaming system performance and provides valuable reference to the experimental results and the solutions.
引文
[1]Akamai [EB/OL]. http://www.akamai.com
[2]S.Gadde,J.Chase and M.Rabinovich.Web caching and content distribution:a view from the interior[A].In Proceedings of 5th International Web Caching and Content Delivery Workshop[C].Lisb on Portugal:Elsevier Science
[3]BitTorrent[EB/OL]. http://bitconjurer.org/BitTorrent,2001-04
[4]Emule[EB/OL]. http://www.emule-project.net.
[5]H. Schulzrinne, S. Casner, R. Frederick, and V. Jacobson.RTP:A Transport Protocol for Real-Time Applications[J].IETF RFC 1889, January 1996.
[6]H. Schulzrinne, A. Rao, and R. Lanphier. Real Time Streaming Protocol (RTSP)[J]. IETF RFC 2326, April 1998.
[7]S. Carter and D. Long. Improving video-on-demand server efficiency through stream tapping[A]. In Computer Communications and Networks,1997. Proceedings., Sixth International Conference[C] on, pages 200-207,22-25 Sept. 1997.
[8]A. Dan, D. Sitaram, and P. Shahabuddin. Scheduling policies for an on-demand video server with batching[A]. In MULTIMEDIA'94:Proceedings of the second ACM international conference on Multimedia[C], pages 15-23, New York, NY, USA,1994. ACM Press
[9]D. Eager, M. Vernon, and J. Zahorjan. Minimizing bandwidth requirements for on-demand data delivery[J].Knowledge and Data Engineering, IEEE Transactions on,13(5):742-757, Sept.-Oct.2001.
[10]L.Golubchik, J.C.S. Lui, and R. Muntz. Reducing i/o demand in video-on-demand storage servers[A]. In SIGMETRICS'95/PERFORMANCE' 95:Proceedings of the 1995 ACM SIGMETRICS joint international conference on Measurement and modeling of computer systems[C], pages 25-36, New York, NY, USA,1995. ACM Press
[11]S. Viswanathan and T. Imielinski. Metropolitan area video-on-demand service using pyramid broadcasting.Multimedia Systems,4(4):197-208, August 1996.
[12]V. Padmanabhan, H. Wang, P. Chou, and K. Sripanidkulchai. Distributing streaming media content using cooperative networking[A]. Proceedings of the 12th international workshop on Network and operating systems support for digital audio and video[C], pages 177-186,2002.
[13]L. de Pinho, E. Ishikawa, and C. de Amorim. GloVE:A distributed environment for scalable video-on-demand systems[A]. Int. J. High Perform. Comput[C]. Appl. (USA),
[14]V. Padmanabhan, H. Wang, P. Chou, and K. Sripanidkulchai.Distributing streaming media content using cooperative networking[A]. Proceedings of the 12th international workshop on Network and operating systems support for digital audio and video[C], pages 177-186,2002.
[15]R. Rejaie and A. Ortega. PALS:Peer-to-Peer Adapative Layered Streaming[A]. In Proceedings of the 13th International Workshop on Network and Operating System Support for Digital Audio and Video[C], pages 153-161, June 2003.
[16]L. Guo, S. Chen, S. Ren, X. Chen, and S. Jiang. PROP:a Scalable and Reliable P2P Assisted Proxy Streaming System[A]. Distributed Computing Systems, 2004. Proceedings.24th International Conference on[C], pages 778-786,2004.
[17]A. Vlavianos, M. Iliofotou, and M. Faloutsos. BiToS:Enhancing BitTorrent for Supporting Streaming Applications[A]. In Proceedings of the 25th IEEE International Conference on Computer Communications (INFOCOM)[C] pages 1-6, April 2006.
[18]Y. Cui, B. Li, and K. Nahrstedt. oStream:asynchronous streaming multicast in application-layer overlay networks[A]. IEEE Journal on Selected Areas in Communications[C],22(1):91-106,2004.
[19]C. Huang, J. Li, and K. Ross. Peer-Assisted VoD:Making Internet Video Distribution Cheap[A]. In:Proceedings of Sixth Internaltional Workshop on Peer-to-Peer Systems[C],2007.
[20]Y.F. Chen et al. When is P2P Technology Beneficial for IPTV Services[A]. In Proceedings of the 17th International Workshop on Network and Operating System Support for Digital Audio and Video[C], May 2007.
[21]C. Dana, D. Li, D. Harrison, and C.-N. Chuah. BASS:BitTorrent assisted streaming system for video-on-demand[A]. IEEE International Workshop on Multimedia Signal Processing (MMSP)[C], October 2005.
[22]VoD[EB/OL].http://baike.baidu.com/view/57502.htm
[23]吕向辰.P2P技术与应用[M],计算机世界,2002：12,13-1
[24]Napster[EB/OL]. http://www.napster.com.
[25]Gnutella [EB/OL]. http://www.gnutella.corn
[26]I. Clarke, et al.Freenet:A distributed anonymous information storage and retrieval system[A]. In:Proc. ICSI Workshop[C]. Los Alamitos, California: IEEE Computer Society Press,2000.
[27]KaZaA[EB/OL].http://www.kazaa.com.
[28]B.Coheta,Bit Torrent protocol pecification[EB/OL], http://www.bitconjurer.org/ BitTorrent/protocol.html, February 2005.
[29]Cohen B. Incentives Build Robustness In BitTorrent[A]. In:1st Workshop on Economics of Peer-to-Peer Systems Berkeley[C],CA, USA,June 2003.
[30]俞嘉地.BitTorrent对等网络文件共享系统关键技术研究[D].上海：上海交通大学.2007.
[31]徐雷鸣,庞博,赵耀,NS与网络模拟[M],人民邮电出版社,北京2003,25-153.
[32]GT-ITM[EB/OL].http://www.cc.gatech.edu/projects/gtitrn/,2008.
[33]E.Zegura, K. Calvert, and S. Bhattacharjee. How to Model an Intemetwork[A], In:Proceedings of IEE inFOCOM'96[C], San Francisco, CA, USA, Mar 1996, v01.2:594-602.
[34]ZHANG,X, LIU.J, U.B AND YUM, T. S. EDONet/CoolStreaming:A data-driven overlay network for live media streaming[A]. In Proceedings of IEEE INFOCOM[C],2005.
[35]S. Tewari and L. Kleinrock. Analytical Model for Bittorrent-based Live Video Streaming[A], In Proceedings of the IEEE NIME 2007 Workshop[C], January 2007.
[36]Tribler[EB/OL].http://www.tribler.org/
[37]Python [EB/OL].http://python.cn/
[38]Yung Ryn Choe, Derek L. Schuff,Jagadeesh M. Dyaberi, and Vijay S. Pai. Improving VoD Server Efficiency with Bittorrent[A], Proceedings of the 15th international conference on Multimedia[C], September 23-28,2007.
[39]R. Fielding, J. Gettys, J. Mogul, H. Frystyk, L. Masinter,P. Leach, and T. Berners-Lee. Hypertext Transfer Protocol HTTP 1.1 [J]. IETF RFC 2616, June 1999.
[2]S.Gadde,J.Chase and M.Rabinovich.Web caching and content distribution:a view from the interior[A].In Proceedings of 5th International Web Caching and Content Delivery Workshop[C].Lisb on Portugal:Elsevier Science
[3]BitTorrent[EB/OL]. http://bitconjurer.org/BitTorrent,2001-04
[4]Emule[EB/OL]. http://www.emule-project.net.
[5]H. Schulzrinne, S. Casner, R. Frederick, and V. Jacobson.RTP:A Transport Protocol for Real-Time Applications[J].IETF RFC 1889, January 1996.
[6]H. Schulzrinne, A. Rao, and R. Lanphier. Real Time Streaming Protocol (RTSP)[J]. IETF RFC 2326, April 1998.
[7]S. Carter and D. Long. Improving video-on-demand server efficiency through stream tapping[A]. In Computer Communications and Networks,1997. Proceedings., Sixth International Conference[C] on, pages 200-207,22-25 Sept. 1997.
[8]A. Dan, D. Sitaram, and P. Shahabuddin. Scheduling policies for an on-demand video server with batching[A]. In MULTIMEDIA'94:Proceedings of the second ACM international conference on Multimedia[C], pages 15-23, New York, NY, USA,1994. ACM Press
[9]D. Eager, M. Vernon, and J. Zahorjan. Minimizing bandwidth requirements for on-demand data delivery[J].Knowledge and Data Engineering, IEEE Transactions on,13(5):742-757, Sept.-Oct.2001.
[10]L.Golubchik, J.C.S. Lui, and R. Muntz. Reducing i/o demand in video-on-demand storage servers[A]. In SIGMETRICS'95/PERFORMANCE' 95:Proceedings of the 1995 ACM SIGMETRICS joint international conference on Measurement and modeling of computer systems[C], pages 25-36, New York, NY, USA,1995. ACM Press
[11]S. Viswanathan and T. Imielinski. Metropolitan area video-on-demand service using pyramid broadcasting.Multimedia Systems,4(4):197-208, August 1996.
[12]V. Padmanabhan, H. Wang, P. Chou, and K. Sripanidkulchai. Distributing streaming media content using cooperative networking[A]. Proceedings of the 12th international workshop on Network and operating systems support for digital audio and video[C], pages 177-186,2002.
[13]L. de Pinho, E. Ishikawa, and C. de Amorim. GloVE:A distributed environment for scalable video-on-demand systems[A]. Int. J. High Perform. Comput[C]. Appl. (USA),
[14]V. Padmanabhan, H. Wang, P. Chou, and K. Sripanidkulchai.Distributing streaming media content using cooperative networking[A]. Proceedings of the 12th international workshop on Network and operating systems support for digital audio and video[C], pages 177-186,2002.
[15]R. Rejaie and A. Ortega. PALS:Peer-to-Peer Adapative Layered Streaming[A]. In Proceedings of the 13th International Workshop on Network and Operating System Support for Digital Audio and Video[C], pages 153-161, June 2003.
[16]L. Guo, S. Chen, S. Ren, X. Chen, and S. Jiang. PROP:a Scalable and Reliable P2P Assisted Proxy Streaming System[A]. Distributed Computing Systems, 2004. Proceedings.24th International Conference on[C], pages 778-786,2004.
[17]A. Vlavianos, M. Iliofotou, and M. Faloutsos. BiToS:Enhancing BitTorrent for Supporting Streaming Applications[A]. In Proceedings of the 25th IEEE International Conference on Computer Communications (INFOCOM)[C] pages 1-6, April 2006.
[18]Y. Cui, B. Li, and K. Nahrstedt. oStream:asynchronous streaming multicast in application-layer overlay networks[A]. IEEE Journal on Selected Areas in Communications[C],22(1):91-106,2004.
[19]C. Huang, J. Li, and K. Ross. Peer-Assisted VoD:Making Internet Video Distribution Cheap[A]. In:Proceedings of Sixth Internaltional Workshop on Peer-to-Peer Systems[C],2007.
[20]Y.F. Chen et al. When is P2P Technology Beneficial for IPTV Services[A]. In Proceedings of the 17th International Workshop on Network and Operating System Support for Digital Audio and Video[C], May 2007.
[21]C. Dana, D. Li, D. Harrison, and C.-N. Chuah. BASS:BitTorrent assisted streaming system for video-on-demand[A]. IEEE International Workshop on Multimedia Signal Processing (MMSP)[C], October 2005.
[22]VoD[EB/OL].http://baike.baidu.com/view/57502.htm
[23]吕向辰.P2P技术与应用[M],计算机世界,2002：12,13-1
[24]Napster[EB/OL]. http://www.napster.com.
[25]Gnutella [EB/OL]. http://www.gnutella.corn
[26]I. Clarke, et al.Freenet:A distributed anonymous information storage and retrieval system[A]. In:Proc. ICSI Workshop[C]. Los Alamitos, California: IEEE Computer Society Press,2000.
[27]KaZaA[EB/OL].http://www.kazaa.com.
[28]B.Coheta,Bit Torrent protocol pecification[EB/OL], http://www.bitconjurer.org/ BitTorrent/protocol.html, February 2005.
[29]Cohen B. Incentives Build Robustness In BitTorrent[A]. In:1st Workshop on Economics of Peer-to-Peer Systems Berkeley[C],CA, USA,June 2003.
[30]俞嘉地.BitTorrent对等网络文件共享系统关键技术研究[D].上海：上海交通大学.2007.
[31]徐雷鸣,庞博,赵耀,NS与网络模拟[M],人民邮电出版社,北京2003,25-153.
[32]GT-ITM[EB/OL].http://www.cc.gatech.edu/projects/gtitrn/,2008.
[33]E.Zegura, K. Calvert, and S. Bhattacharjee. How to Model an Intemetwork[A], In:Proceedings of IEE inFOCOM'96[C], San Francisco, CA, USA, Mar 1996, v01.2:594-602.
[34]ZHANG,X, LIU.J, U.B AND YUM, T. S. EDONet/CoolStreaming:A data-driven overlay network for live media streaming[A]. In Proceedings of IEEE INFOCOM[C],2005.
[35]S. Tewari and L. Kleinrock. Analytical Model for Bittorrent-based Live Video Streaming[A], In Proceedings of the IEEE NIME 2007 Workshop[C], January 2007.
[36]Tribler[EB/OL].http://www.tribler.org/
[37]Python [EB/OL].http://python.cn/
[38]Yung Ryn Choe, Derek L. Schuff,Jagadeesh M. Dyaberi, and Vijay S. Pai. Improving VoD Server Efficiency with Bittorrent[A], Proceedings of the 15th international conference on Multimedia[C], September 23-28,2007.
[39]R. Fielding, J. Gettys, J. Mogul, H. Frystyk, L. Masinter,P. Leach, and T. Berners-Lee. Hypertext Transfer Protocol HTTP 1.1 [J]. IETF RFC 2616, June 1999.