PhD Proposal: Yousuk Seung, March 18, 2 p.m., ACES 6.336

Contact Name: 
Lydia Griffith
Date: 
Mar 18, 2013 2:00pm - 4:00pm

PhD Proposal: Yousuk Seung

Date & Time : 2-4pm, 3/18 Mon
Place : ACES 6.336
Research Supervisor:  Yin Zhang

Title : Optimizing Mobile Multimedia Content Delivery

Abstract :

Demand for high quality multimedia data over the Internet in the
mobile environment has been growing rapidly. Popularity of video streaming
has greatly increased recently and real-time entertainment data including video
and audio now accounts for a majority of Internet traffic. Video conferencing
is also gaining popularity and is becoming accessible to non-enterprise users.
With the advent of mobile Internet such multimedia applications are now
becoming more popular in the mobile environment, however efficient content
delivery in such cases yet remains a challenging problem.

The mobile environment is significantly more unfriendly than the wired
environment in many ways for multimedia applications. Network resources are
more limited and expensive and the condition is not as stable which makes
prediction difficult. However multimedia applications are generally delay intol-
erant and bandwidth demanding, and with users moving their demand could be
much more dynamic and harder to anticipate. Due to such reasons many exist-
ing mobile applications for multimedia services such as multimedia streaming
or video conferencing are not well optimized for mobile environment and show
unsatisfactory performance.

We target three multimedia content delivery applications and optimize
with limited and unpredictable network conditions typical in the mobile Inter-
net environment.

Vehicular Networks: Vehicular networks have emerged from the strong
desire to communicate on the move. We explore the potential of supporting
high-bandwidth applications such as video streaming in vehicular networks.
Challenges include limited and expensive cellular network, etc.

Video Conferencing: The popularity of Internet video conferencing has
surged over the past few years, but supporting high-quality large video confer-
ences at a low cost remains a significant challenge due to stringent performance
requirements, limited and heterogeneous client. We develop a simple yet ef-
fective valiant multicast routing to select application-layer routes and adapt
streaming rates according to dynamically changing network condition in a
swift and lightweight way enough to be implemented on mobile devices.

Mobile Video Streaming and Rate Adaptation : We target the rate ada-
pative online video streaming application and are developing a real system that
implements a novel rate selection strategy that can maximize user-perceived
video quality. The system aims to minimize rate changes while maximizing
average video streaming rate by dynamically adapting to changing network
conditions. This is an onoing work and we plan to further extend the idea to
optimize aggregated user-perceived streaming quality of clients that share the
same network bottleneck through inter-client optimization.