Qoe Research On Dynamic Adaptive Streaming Over Http(Dash)

With the rapid development of wireless communications and the Internet, as well as the explosive popularization of Android platform, iPad, netbooks and other terminal devices, it brings huge burden for the network to meet users’ increasing video demand. In order to improve such situation, the adaptive streaming is invented, and DASH (Dynamic Adaptive Streaming over HTTP) turns out the first international standard on adaptive streaming, and has been adopted by many SPs. However, the QoE (Quality of Experience) research about adaptive streaming is still rare, not to mention about DASH. Therefore, this article is dedicated to the QoE research about DASH service, evaluating its QoE performance, exploring its unique QoE characteristics, and building a new QoE evaluation methodology that is applicable to DASH service.Firstly, the article introduces the prospect of DASH from its developing process, and gives a detailed description about MPEGDASH standard, and its QoE expectations. Secondly, it’s evaluated to assess the QoE performance for DASH service through the controlled subjective tests, and to explore its new features. Finally, deeper research on "bitrate distribution" is conducted to explore its psychological mechanism, and then a new evaluation methodology is established to realize the QoE evaluation for DASH service with the help of divide-and-conquer strategy. It’s proven by the subjective tests that there is great promotion on the accuracy and the sensitivity to "bitrates distribution" of new evaluation methodology.