基于Q-learning的HTTP自适应流码率控制方法研究
发布时间：2019-01-21 11:30
【摘要】：基于HTTP的自适应流HAS已经成为自适应视频流服务的标准。在HAS客户端网络状态多变的情况下,硬编码形式的码率决策方法灵活性偏低,对用户体验考虑不足。为了优化用户体验质量(Qo E),提出一种基于Q-Learning的码率控制算法,结合HTTP自适应视频流客户端环境进行建模并定义状态转移规则;量化与用户Qo E相关的参数,构建新的回报函数;实验表明引入Q-Learning进行码率调整的自适应算法在码率切换的稳定性方面表现较好。
[Abstract]:Adaptive stream HAS based on HTTP has become the standard of adaptive video stream service. When the state of HAS client network is changeable, the rate decision method of hard coding is less flexible, and the user experience is not considered enough. In order to optimize user experience quality (Qo E), a rate control algorithm based on Q-Learning is proposed, which is based on HTTP adaptive video stream client environment and defines state transition rules. The parameters related to the user Qo E are quantified and a new return function is constructed, and the experiments show that the adaptive algorithm with Q-Learning for rate adjustment performs well in the stability of bit rate switching.
【作者单位】： 浙江工业大学计算机科学与技术学院;
【基金】：浙江省重大科技专项重大工业基金资助项目(No.2012C11026-2) 杭州市重大科技创新专项基金资助项目(No.20132011A16)~~
【分类号】：TN919.8
[Abstract]:Adaptive stream HAS based on HTTP has become the standard of adaptive video stream service. When the state of HAS client network is changeable, the rate decision method of hard coding is less flexible, and the user experience is not considered enough. In order to optimize user experience quality (Qo E), a rate control algorithm based on Q-Learning is proposed, which is based on HTTP adaptive video stream client environment and defines state transition rules. The parameters related to the user Qo E are quantified and a new return function is constructed, and the experiments show that the adaptive algorithm with Q-Learning for rate adjustment performs well in the stability of bit rate switching.
【作者单位】： 浙江工业大学计算机科学与技术学院;
【基金】：浙江省重大科技专项重大工业基金资助项目(No.2012C11026-2) 杭州市重大科技创新专项基金资助项目(No.20132011A16)~~
【分类号】：TN919.8
【相似文献】
相关期刊论文 前10条
1 徐荣杰,杨广文;一种带场景变换检测的码率控制策略[J];计算机工程与应用;2003年28期
2 王红玉;管鑫;周洪敏;;基于场景切换的码率控制方法[J];科技资讯;2008年03期
3 朱仲杰;白永强;蒋刚毅;粱丰;;基于双模预测的精确码率控制[J];中国图象图形学报;2010年04期
4 董焱鑫,李桂苓;模块化联合码率控制技术[J];电子技术应用;2000年08期
5 范晓东,余松煜,孙军;基于场景切换检测的码率控制策略[J];上海交通大学学报;2000年06期
6 环，
本文编号：2412621
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2412621.html
