首页 | 本学科首页   官方微博 | 高级检索  
     

基于多步下溢概率估计的自适应媒体播放算法
引用本文:张辉帅,王劲林,朱小勇,孙鹏.基于多步下溢概率估计的自适应媒体播放算法[J].微计算机应用,2012,1(3):19-23.
作者姓名:张辉帅  王劲林  朱小勇  孙鹏
作者单位:1. 中国科学院研究生院,北京100039;中科院声学研究所国家网络新媒体工程技术研究中心,北京100190
2. 中科院声学研究所国家网络新媒体工程技术研究中心,北京,100190
基金项目:863课题,科技支撑计划课题,中国科学院战略性先导科技专项子课题
摘    要:自适应播放(AMP)是一种通过动态调整播放速率减少这种中断的技术。多数AMP算法都基于缓存的满溢度或者其变化调整播放速率,其难点在于如何选择合适的调整门限值。本文算法,利用滑动窗口估计到达帧率的经验分布,再利用统计计算方法估计多步缓存下溢概率,然后根据一步和多步下溢概率调整播放速率。仿真实验证明在一般的信道假设和多种片源的测试中,本算法较其他同类算法有更好的表现。

关 键 词:自适应媒体播放  媒体缓存  视频流化  概率估计

Adaptive Media Playout Based on Multi- step Underflow Probability
ZHANG Huishuai , WANG Jinlin , ZHU Xiaoyong , SUN Peng.Adaptive Media Playout Based on Multi- step Underflow Probability[J].Microcomputer Applications,2012,1(3):19-23.
Authors:ZHANG Huishuai  WANG Jinlin  ZHU Xiaoyong  SUN Peng
Affiliation:1 Graduate University, Chinese Academy of Sciences, Beijing, 100039, China;2 National Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences, Beijing, 100190, China)
Abstract:Adaptive Media Playout(AMP) is an effective method to reduce the playout interruptions of the video streaming applications. Most proposed AMP algorithms trigger the adjustment of playback speed based on the buffer' s fullness or its variations. Where the conundrum is hou to select the appropriate threshold. Our AMP algorithm firstly estimates the empirical distribution of arrival frame rate in sliding windows and then works out the one - step and multi - step underflow probabilities using statistical method. Finally, the play- back speed is altered according to the underflow probabilities. Experiments with various encoded video sources and general VBR channels show our AMP algorithm outperforms other congeneric algorithms.
Keywords:Adaptive media playout  media buffering  video streaming  probability estimation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号