基于最佳分段点估计的流媒体非均匀分段方法

doi:10.11959/j.issn.1000-0801.2015232

摘要/Abstract

摘要：

摘要：针对命中率随存储的流媒体片段流行度变化的特征，提出了一种新的基于最佳分段点估计的流媒体非均匀分段方法，根据不同存储大小下的流媒体外部分界流行度对其内部最佳分段点进行估计，进而把每个视频分成高流行度段和低流行度段两个片段。实验结果表明，与均分分段相比，该方法能减少流媒体的片段数，提高缓存命中率。

关键词: 流行度, 分段, 流媒体

Abstract:

According to the relationship between the byte hit rates and the popularities of the stored streaming media segments, a new non-uniform segmentation method for the streaming media based on best segmentation point estimation was proposed. In the proposed method, the best internal segmentation points of the streaming media were estimated according to the external boundary popularity of the streaming media at the current storage size of the proxy. Then each video was divided into high popularity and low popularity parts using the estimated best internal segmentation point. Experimental results show that the proposed method can effectively reduce the number of segments and improve byte hit rates comparing with the uniform segmentation method.

Key words: popularity, segmentation, streaming media

邓宇珊,庄一嵘,陈戈,张军. 基于最佳分段点估计的流媒体非均匀分段方法[J]. 电信科学, 2015, 31(9): 124-129.

Yushan Deng,Yirong Zhuang,Ge Chen,Jun Zhang. Non-Uniform Segmentation Method for Streaming Media Based on Best Segmentation Point Estimation[J]. Telecommunications Science, 2015, 31(9): 124-129.

图/表 6

图1

图2

图3

图4

图5

表2

参考文献 17

1	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．软件学报， 2014,25（1）:98～117 Zhang Y J , He M , Meng X W . Research on CDN-P2P system over user requirements. Journal of Software, 2014,25（1）:98～117
2	Theodorou M , Paterakis M . Efficient cache management policiesfor streaming proxies supporting complete and partial video playback. Proceedings of the 5th International Conference on Next Generation Networks and Services （NGNS 2014）, Casablanca, Morocco, 2014:270～277
3	何智聪，谷光昭，王新等．基于可重构路由器上缓存的流媒体协作分发策略．通信学报， 2012,33（6）:82～90 He Z C , Gu G Z , Wang X , et al. Novel cooperative caching strategies for video streaming distribution based on reconfiguration routers. Journal on Communications, 2012,33（6）:82～90
4	王蒙蒙．基于异构网络的交互式流媒体缓存替换算法．信息通信， 2013（6）:71～72 Wang M M . Cache replacement algorithm based on heterogeneous network of interactive streaming media file. Information and Communications, 2013（6）:71～72
5	Sen S , Rexford J , Towsley D . Proxy prefix caching for multimedia streams. Proceedings of IEEE Conference On Computer Communications, New York, USA, 1999:1310～1319
6	Chen S Q , Shen B , Wee S , et al. Adaptive and lazy segmentation based proxy caching for streaming media delivery. Proceedings of the 13th International Workshop on Network and Operating Systems Support for Digital Audio and Video, Monterey, USA, 2003:22～31
7	Wu K L , Yu P S , Wolf J L . Segment-based proxy caching of multimedia streams. Proceedings of the 10th International Conference on World Wide Web, Hong Kong, China, 2001:36～44
8	余江，杨宗凯，杜旭等．基于两点流行度的流媒体缓存算法．华中科技大学学报（自然科学版）， 2006,34（10）:15～17 Yu J , Yang Z K , Du X , et al. Two-point popularity-based caching algorithm for streaming media. Journal of Huazhong University of Science and Technology（Nature Science）, 2006,34（10）:15～17
9	邓宇珊，梁洁，张军等．基于流行度分类的流媒体分段策略．电信科学， 2015,31（1）:152～157 Deng Y S , Liang J , Zhang J , et al. Segmentation strategy of streaming media based on popularity classifying. Telecommunications Science, 2015,31（1）:152～157
10	杨传栋，余镇危，王行刚等．基于流行度预测的流媒体代理缓存替换算法．计算机工程， 2007,33（7）:99～100,129 Yang C D , Yu Z W , Wang X G , et al. Proxy cache replacement algorithm based on popularity prediction of streaming media file. Computer Engineering, 2007,33（7）:99～100,129
11	张艳，牛朵朵．基于最小代价的流媒体缓存替换算法研究．中原工学院学报， 2012,23（5）:73～75 Zhang Y , Niu D D . Research of caching replacement algorithm of streaming media based on minimize cost. Journal of Zhongyuan Institute of Technology, 2012,23（5）:73～75
12	Yu J , Chun T C , Yang Z K , et al. A proxy caching algorithm based on user access preference in streaming media. Microelectronics ＆ Computer, 2006,23（11）:19～22,25
13	刘宜宁，赵正德，全卫新等．基于媒体流行度和前缀缓存的缓存替换算法．中国图像图形学报， 2007,12（10）:1753～1756 Liu Y N , Zhao Z D , Quan W X , et al. Proxy cache replacement algorithm based on popularity and prefix caching. Journal of Image and Graphics, 2007,12（10）:1753～1756
14	周轴．视频点播系统访问行为研究：测量、分析与建模（博士学位论文）．合肥：中国科学技术大学， 2009 Zhou Z . Research on the access behavior of video on demand system: measurement, analysis and modeling （doctor dissertation）. Hefei: University of Science and Technology of China, 2009
15	余江．流媒体代理缓存算法研究（博士学位论文）．武汉：华中科技大学， 2006 Yu J . Research on proxy caching algorithms for streaming media （doctor dissertation）. Wuhan: Huazhong University of Science and Technology, 2006
16	李晓楠，朱永宽，张来胜．基于用户行为的流媒体分段缓存算法．中原工学院学报， 2010,21（4）:62～64,69 Li X N , Zhu Y K , Zhang L S . Segmentation caching algorithm of streaming media based on user accessing. Journal of Zhongyuan Institute of Technology, 2010,21（4）:62～64,69
17	刘磊，熊小鹏．最小驻留价值缓存替换算法．计算机应用， 2013,33（4）:1018～1022 Liu L , Xiong X P . Least cache value replacement algorithm. Journal of Computer Applications, 2013,33（4）:1018～1022

存储占比	均匀分段总分段数	最佳分段点估计分段策略总分段数	减少量百分比
10%	2 400	562	77%
20%	2 800	745	73%
30%	3 200	785	75%
40%	4 800	780	84%
50%	2 400	778	68%
60%	2 800	767	73%
70%	8 800	748	92%
80%	1 600	726	55%
90%	6 800	683	90%

[1]	马学森, 陈树友, 许向东, 储昭坤. 基于神经网络与马尔可夫组合模型的视频流行度预测算法[J]. 电信科学, 2021, 37(8): 18-26.
[2]	蔡秉江. 基于LTE资源块感知的自适应无线流媒体系统[J]. 电信科学, 2021, 37(1): 147-158.
[3]	王博伟,谭劲. 基于两次分段弱选择的压缩感知子空间追踪算法[J]. 电信科学, 2020, 36(5): 83-92.
[4]	钟其柱. 基于熵值和模糊层次法评估媒体流的研究及应用[J]. 电信科学, 2020, 36(4): 145-153.
[5]	唐永林,刘锋. SRv6的分段标识符结构的改善[J]. 电信科学, 2020, 36(4): 170-174.
[6]	张雪垣,王让定,严迪群,林昱臻. 基于分段STC的音频隐写算法[J]. 电信科学, 2019, 35(7): 115-123.
[7]	倪妍红,楼向雄,包建荣. 改进的分段CRC辅助的删余Polar译码[J]. 电信科学, 2019, 35(3): 91-98.
[8]	曹畅, 张帅, 唐雄燕. 下一代智能融合城域网方案[J]. 电信科学, 2019, 35(10): 51-59.
[9]	沈周青,尚俊娜. 极化码的多比特译码算法研究[J]. 电信科学, 2018, 34(11): 77-86.
[10]	曲桦,李岩松,刘军,赵季红,郭涯. 一种基于SDN的CCN集中控制缓存决策方法[J]. 电信科学, 2017, 33(5): 12-20.
[11]	高晓敏,刘雪晶,李佳伟,罗洪斌,董平. SR软件路由器的设计与实现[J]. 电信科学, 2017, 33(2): 104-112.
[12]	何晓明,卢泉,邢亮. 分段路由网络研究及其在流量工程中的应用[J]. 电信科学, 2016, 32(6): 186-194.
[13]	祝谷乔,宋皓. MPEG-DASH与HLS流传输技术的比较分析[J]. 电信科学, 2015, 31(4): 22-26.
[14]	陈皇卿,钟晓峰,孙剑,王京. CASoRT系统中基于聚集特性的在线流行度预测方法[J]. 电信科学, 2015, 31(11): 1-6.
[15]	邓宇珊,陈戈,张军,梁洁. 基于流行度分类的流媒体分段策略[J]. 电信科学, 2015, 31(1): 2015034-.