YouTube Can Do Better: Getting the Most Out of Video Adaptation
QCMan 2016 (QCMan 2016), Würzburg, Germany, (September 2016)Abstract
In this paper, we look at YouTube's algorithm for the adaptation of the
video resolution, and analyze how much more mileage can be gotten out of
it. In previous work, we analyzed the efficiency of YouTube's adaptation
strategy by using isotonic regression that has the disadvantage of not
eliminate stalling events. As a next step, we apply use an optimization
problem to provide the optimal result in any given scenario while avoiding
stalling events. This paper provides an outlook on the potential that is
left for improving YouTube's adaptation mechanism that is essential to
managing the QoE in video streaming.
