最大紧迫度优先的流媒体批调度算法
doi: 10.3724/SP.J.1146.2007.01993
The Maximum Urgency First Batching Algorithm for Streaming Media
-
摘要: 批处理是一种简单有效的流媒体调度机制。通过把批处理调度归结为平均访问延迟最优化问题,得到影响批处理算法性能的关键因素。基于分析结果,本文提出一种最大紧迫度优先的批处理机制及其实现方案。紧迫度的定义考虑了等待队列的多个属性。一旦有空闲信道时,该机制选择具有最大紧迫度的视频队列进行服务。详细的仿真结果证明了所提方案的有效性和可行性。Abstract: Batching is a simple but efficient scheduling mechanism for streaming media. An optimization model for average access delay is formulated to obtain the key aspects affecting the performance of Batching. Based on the analytical results, the maximum urgency first batching policy and its heuristic implementations are proposed. The definition of urgency considers multiple attributes of the waiting queue. Whenever a channel becomes available, the proposed policy schedules the video with the largest urgency value. The effectiveness and feasibility of the presented schemes are evaluated and justified experimentally using simulation.
-
[1] Huang C, Li J, and Ross K W. Can Internetvideo-on-demand be profitable? Proc. of ACM SIGCOMM,Kyoto, Japan, August 2007: 133-144. [2] Dan A, Sitaram D, and Shahabuddin P. Scheduling policiesfor an on-demand video server with batching. Proc. of SecondACM Multimedia Conf., San Francisco, USA, October 1994:391-398. [3] Aggarwal C C, Wolf J L, and Yu P S. The maximum factorqueue length batching scheme for video-on-demand systems[J].IEEE Trans. on Comput.2001, 50(2):97-110 [4] Fonseca N L S and Facanha R A. The look-aheadmaximize-batch batching policy[J].IEEE Trans. on Multimedia.2002, 4(1):114-120 [5] Cao W and Aksoy D. MAI: Multiple attributes integrationfor deadline aware broadcast scheduling. Proc. of IEEE IntlSymposium on Computer Networks (ISCN), Istanbul,Turkey, June 2006: 55-61. [6] Sarhan N J and Das C R. A new class of scheduling policiesfor providing time of service guarantees in video-on-demandservers. Proc. of 7th IFIP/IEEE Intl Conf. on Managementof Multimedia Networks and Services (MMNS), California,USA, October 2004: 127-139. [7] Alsmirat M A, Al-Hadrusi M, and Sarhan N J. Analysis ofwaiting-time predictability in scalable media streaming. Proc.of ACM Mulitmedia, Augsburg, Germany, September 2007:727-736. [8] Sarhan N J and Qudah B. Efficient cost-based scheduling forscalable media streaming. Proc. of Multimedia Computingand Networking Conf. (MMCN), California, USA, January2007: 134-145. [9] 蔡青松, 李子木, 胡建平. Internet上的流媒体特性及用户访问行为研究. 北京航空航天大学学报, 2005, 31(1): 25-30.Cai Qing-song, Li Zi-mu, and Hu Jian-ping. Study oncharacterizing streaming media and user access to media onthe Internet. Journal of Beijing University of Aeronautics andAstronautics, 2005, 31(1): 25-30. [10] Yu H L, Zheng D D, and Zhao B Y, et al.. Understanding userbehavior in large scale video-on-demand systems. Proc. ofEuroSys, Leuven, Belgium, April 2006: 333-344. [11] Guo L, Tan E, and Chen S, et al.. Does Internet media trafficreally follow zipf-like distribution? Proc. of ACMSIGMETRICS, California, USA, June 2007: 359-360.
点击查看大图
计量
- 文章访问数: 3103
- HTML全文浏览量: 80
- PDF下载量: 919
- 被引次数: 0