Abstract
The transmission of video data is a major part of traffc on
today's Internet. Since the Internet is a highly dynamic environment,
quality adaptation is essential in matching user device resources with
the streamed video quality. This can be achieved by applying mechanisms
that follow the Scalable Video Coding (SVC) standard, which
enables scalability of the video quality in multiple dimensions. In SVCbased
streaming, adaptation decisions have long been driven by Quality
of Service (QoS) metrics, such as throughput. However, these metrics do
not very well match the way human users perceive video quality. Therefore,
in this paper, the classical SVC-based video streaming approach is
expanded to consider Quality of Experience (QoE) for adaptation decisions.
The video quality is assessed using existing objective techniques
with a high correlation to the human perception of video quality. The approach
is evaluated in context of a P2P-based Video-on-Demand (VoD)
system and shows that by making peers favor always layers with a high
estimated QoE but not necessarily high bandwidth requirements, the
performance of the entire system can be enhanced in terms of playback
delay and SVC video quality by up to 20%. At the same time, content
providers are able to reduce up to 60% of their server costs, compared
to the classical QoS-based approach.
Users
Please
log in to take part in the discussion (add own reviews or comments).