计算机科学 ›› 2015, Vol. 42 ›› Issue (7): 258-261.doi: 10.11896/j.issn.1002-137X.2015.07.055
沈项军,姚 银,查正军
SHEN Xiang-jun, YAO Yin and ZHA Zheng-jun
摘要: 在P2P(Peer to Peer)网络视频点播中设计合理的视频内容预取策略,可以有效地解决视频观看中出现的缓冲等待时间过长、服务器负载过大等问题。而现有的P2P网络视频预取方法往往考虑了视频内容的发现而忽略了节点本身的状态监控,在网络拥塞发生时存在视频播放观看不佳的情况。提出了一种基于拥塞发现的强化学习P2P网络视频点播预取策略,通过监测节点的拥塞状况以及带宽等参数,采用Q-learning学习算法综合评估网络节点,引导视频预取节点的选择,减少了对网络拥塞节点视频内容的预取。实验结果表明,该方法能提高视频播放流畅性,避免用户在节点发生拥塞时等待时间过长,提高了视频播放效率。
[1] Rao L.comScore:YouTube Reaches All-Time High of 14.6 Billion Videos Viewed In May.http://techcrunch.com/2010/06/24/comscore-youtube-reaches-all-time-high-of-14-6-billion-videos-viewed-in-may/,2010 [2] Cisco I.Cisco Visual Networking Index:Forecast and Metho-dology[S].Cisco System,2008 [3] Passarella A.A survey on content-centric technologies for thecurrent Internet CDN and P2P solutions[J].Computer Communications,2012,5(1):1-32 [4] Thampi S M,Sekaran C K.Survey of search and replicationschemes in unstructured P2P networks[J].Network Protocols and Algorithms,2010,2(1):93-131 [5] Thampi S M,Sekaran C K.Q-learning Based CollaborativeLoad Balancing Using Distributed Search for Unstructured P2P Networks[C]∥33rd IEEE Conference on Local Computer Networks,2008:797-802 [6] Bergamini J A,Haungs M.Enabing P2P Coorperation WMSproxy Caching and Prefetching in an Educational Environment[C]∥ 10th AGILE International Conference on Geographic Information Science.2007:1-14 [7] Ahmed T,Abbasi U,Medjiah S.Efficient IPTV Services Deli-very using SVC Adaptation and Cooperative Prefetching[J].IEEE Comsoc Mmtc E-Letter,2011,6(1):49-50 [8] Guo Hong-fang,Lin Yu-song,Wang Zong-min.User-behaviorbased self-adaptive indexing for VoD service[J].Computer Applications Research,2012,9(4):86(下转第275页)(上接第261页) [9] Yu Hong-liang,Zheng Dong-dong,Zhao B Y,et al.UnderstandingUser Behavior in Large-scale Video Streaming Services[J].Computer Science Department,2006,0(4):18-21 [10] Riad A M,Elmogy M,Schehab A I.A Framework for Cloud P2P VoD System based on User’s Behavior Analysis[J].International Journal of Computer Applications,2013,6(6):20-26 [11] He Yi-feng,Shen Guo-bin,Xiong Yong-qiang.Optimal Prefet-ching Scheme in P2P VoD Applications With Guided Seeks[J].IEEE Transaction on Multimedia,2009,1(2):138-151 [12] 王庆波,代亚非,田敬,等.基于特征信息定位的P2P网络模型:Barnet[J].软件学报,2003,4(8):1481-1488 Wang Qing-bo,Dai Ya-fei,Tian Jing,et al.An Infrastructure for Attribute Addressable P2P Network:Barnet[J].Jornal of SoftWare,2003,4(8):1481-1488 [13] 张玉洁,何明,孟祥武.基于用户需求的内容分发点对点网络系统研究[J].软件学报,2014,5(1):98-117 Zhang Yu-jie,He Ming,Meng Xiang-wu.Research on CDN-P2P system over user requirements[J].Journal of Software,2014,5(1):98-117 [14] Sutton R S,Barto A G.Reinforcement Learning-An Introduction[M].MIT Press,1998 [15] Tanner B,Sutton R S.Temporal-Difference Networks with Eligibility Traces[C]∥Proceedings of the 22nd International Conference on Machine Learning.2005:1313-1320 [16] Murphy S A.A Generalization Error for Q-Learning[J].Journal of Machine Learning Research,2005,6:1073-1097 [17] Grondman I,Busoniu L,Lopes G A D,et al.A Survey of Actor-Critic Reinforcement Learning:Stan-dard and Natural Policy Gradients[J].IEEE Transactions on Systems,2012,2(6):1291-1307 [18] Mahadevan S,Maggioni M.Proto-value Functions:A Laplacian Framework for Learning Representation and Control in Markov Decision Processes[J].Journal of Machine Learning Research,2007,8(10):2169-2231 [19] Kwong K W,Tsang D H K.A Congestion_Aware Search Protocol for Unstructured Peer-to-Peer Networks[C]∥ Parallel and Distributed Processing and Applications:Second International Symposium(ISPA).HongKong,China,2005:319-329 [20] Huang Cheng,Li Jin,Ross K W.Can Internet Video-on-Demand be Profitable?[J].ACM SIGCOMM Computer Communication Review,2007,7(4):133-144 [21] Abbasi U,Ahmed T.COOCHING:Cooperative PrefetchingStrategy for P2P Video-on-Demand System[C]∥Wired-Wireless Multimedia Networks and Services Management:IEEE International Conference on Management of Multimedia and Mobile Networks and Service(MIMNS 2009).2009:195-200 |
No related articles found! |
|