Study of user QoE improvement for dynamic adaptive streaming over HTTP (MPEG-DASH)

摘要

Video streaming over HTTP is becoming the de facto dominating paradigm for today's video applications. HTTP as an over-the-top (OTT) protocol has been leveraged for quality video traversal over the Internet. High user-received quality-of-experience (QoE) is driven not only by the new technology, but also by a wide range of user demands. Given the limitation of a traditional TCP/IP network for supporting video transmission, the typical on-off transfer pattern is inevitable. Dynamic adaptive streaming over HTTP (DASH) establishes a simple architecture and enables new video applications to fully utilize the exiting physical network infrastructure. By deploying robust adaptive algorithms at the client side, DASH can provide a smooth streaming experience. We propose a dynamic adaptive algorithm in order to keep a high QoE for the average user's experience. We formulated our QoE optimization in a set of key factors. The results obtained by our empirical network traces show that our approach not only achieves a high average QoE but it also works stably under different network conditions.