Evaluation of CMAF in live streaming scenarios
Evaluation of CMAF in live streaming scenarios
复制标题
CMAF在直播场景下的评估
DOI：
--
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
S. Appleby
中科院分区：
文献类型：
--
作者：
Tomasz Lyko;M. Broadbent;N. Race;M. Nilsson;Paul Farrow;S. Appleby
HTTP自适应流传输（HAS）技术（诸如MPEG DASH）现在被广泛用于向大量观众递送电视服务。在HAS中，客户端使用HTTP请求内容片段，其中ABR算法选择请求每个片段的质量以在避免停顿的情况下折衷视频质量。与传统广播相比，这引入了显著的端到端延迟，这是由于客户端需要足够大的缓冲器以使ABR算法及时地对网络条件的变化做出反应。最近标准化的公共媒体应用格式（CMAF）通过将段定义为由独立可传输的块组成来帮助解决延迟问题。在本文中，我们描述了一个模拟模型，我们已经开发了四个流行的ABR算法，使用DASH和CMAF在各种低延迟直播流媒体场景的性能进行评估。实际网络条件用于评估，其基于从商业直播电视服务的CDN日志获取的吞吐量数据。我们量化的ABR算法的性能，使用选择的QoE指标，并表明CMAF可以显着提高ABR的性能在低延迟的情况下。
HTTP Adaptive Streaming (HAS) technologies such as MPEG DASH are now used extensively to deliver television services to large numbers of viewers. In HAS, the client requests segments of content using HTTP, with an ABR algorithm selecting the quality at which to request each segment to trade-off video quality with the avoidance of stalling. This introduces significant end to end latency compared to traditional broadcast, due to the the client requiring a large enough buffer for the ABR algorithm to react to changes in network conditions in a timely manner. The recently standardised Common Media Application Format (CMAF) has helped address the issue of latency by defining segments as composed of independently transferable chunks. In this paper, we describe a simulation model we have developed to evaluate the performance of four popular ABR algorithms using DASH and CMAF in various low latency live streaming scenarios. Realistic network conditions are used for the evaluation, which are based on throughput data taken from the CDN logs of a commercial live TV service. We quantify the performance of the ABR algorithms using a selection of QoE metrics, and show that CMAF can significantly improve ABR performance in low delay scenarios.