Institute of Information Science, Academia Sinica

Today's commercial video streaming services use dynamic rate selection to provide a high-quality user experience. Most services host content on standard HTTP servers in CDNs, so rate selection must occur at the client. We measure three popular video streaming services – Hulu, Netflix, and Vudu – and find that accurate client-side bandwidth estimation above the HTTP layer is hard. As a result, rate selection based on inaccurate estimates can trigger a feedback loop, leading to undesirably variable and low-quality video. We call this phenomenon the downward spiral effect, and we measured it on all three services, present insights into its root causes, and validate initial solutions to prevent it. At least one major video streaming service changed its algorithm due to this work. The resulted algorithm improved video rate, yet could lead to unnecessary rebuffer events.This talk argues that we should do away with estimating network capacity, and instead directly observe and control the playback buffer. We present a class of rate selection algorithms that allow us to optimize the delivered video quality while provably never unnecessarily rebuffering. Our algorithms work with discrete video rates, video chunking and for both CBR and VBR video codecs. This work is awarded IETF/IRTF Applied Networking Research Prize in 2013.

Te-Yuan is currently a Ph.D. candidate in Computer Science department in Stanford University, working with Prof. Nick McKoewn and Prof. Ramesh Johari. She is generally interested in client-side network stack design and multimedia networking. Before joining Stanford, she received her M.S. from National Taiwan University in 2008 and her B.S. from National Chiao-Tung University in 2006. Te-Yuan is a recipient of Stanford Graduate Fellowship (2008-2012), Google Fellowship (2012-2014), and IETF/IRTF Applied Networking Research Prize, 2013.