摘要
为了解决HTTP自适应流中分段较长时带宽估计不准的问题,提出一种改进的带宽估计方法。该方法包括段内估测和段间估测:段内估测通过指数平均方法平滑段内抽样带宽序列求得该段的即时带宽;段间估测通过逻辑斯谛方程预测即时带宽的变化值来求得下一段的估测带宽。在实际的HTTP Live Streaming(HLS)系统上对提出方法进行验证,实验结果表明,该方法不但能有效地捕捉到实际带宽的快速变化,迅速地跟随大的网络波动,而且能够较好地平滑小的网络波动。此外,该方法仅需在应用层做控制,实现简单。
In order to solve the problem of inaccurate bandwidth estimation when the segment duration is relatively long in HTTP adaptive streaming, an improved bandwidth estimation method is proposed, including the intra-segment and inter-segment estimation. The intra-segment estimation smooths the sampled bandwidth sequences in the segment by exponential averaging method, and the instantaneous bandwidth of this segment is obtained. The inter-segment estimation predicts the variation value of instantaneous bandwidth by logistic functions, and the estimated bandwidth of the next segment is obtained. The proposed method is verified through the real HTTP live streaming （HLS） system. Experimental results show that the method can not only effectively track the fast fluctuations of actual bandwidth and rapidly follow the large network fluctuations, but also smooth the small network fluctuations. Furthermore, the method can be easily implemented in the application layer.
作者
耿立宏
朱小勇
郭志川
潘梁
Geng Lihong Zhu Xiaoyong Guo Zhichuan Pan Liang(Vational Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China University of Chinese Academy of Sciences, Beijing 100190, China)
出处
《计算机应用与软件》
2017年第3期109-113,共5页
Computer Applications and Software
基金
国家高技术研究发展计划项目(2015AA015802)
关键词
HTTP
自适应流
带宽估计
指数平均
逻辑斯谛方程
HTTP Adaptive streaming Bandwidth estimation Exponential average Logistic function
