Events Calendar

PhD Preliminary Oral Exam – Kevin Gatimu

qMDP: DASH Adaptation using Queueing Theory within a Markov Decision Process

As video consumption continues to rise rapidly, the Internet landscape is shifting accordingly, becoming increasingly video-dominated. HTTP has become the de facto mechanism for video delivery due to its ubiquity. As a result, Dynamic Adaptive Streaming over HTTP (DASH) has emerged as the open source standard for state-of-the-art in video streaming. In DASH, a client makes HTTP requests to a server for video chunks from a selection of different quality versions. The selected quality is based on the client's perceived bandwidth through a dynamic process called bitrate adaptation. Most bitrate adaptation algorithms tend to have significant ad hoc components, which has led to reinforcement learning (RL) algorithms. However, RL algorithms for DASH tend to trade off between accuracy and efficiency. This work proposes a queueing-theory-based Markov Decision Process (qMDP), which features a simplified state characterization based on an M/D/1/K queueing model. qMDP ! aims to converge towards and optimal solution faster than pre-existing RL methods. This would make it well-suited for online deployment in DASH clients.

Major Advisor: Ben Lee
Committee: Roger Traylor
Committee: Thinh Nguyen
Committee: Huaping Liu
GCR: Janet Nishihara

Friday, November 30, 2018 at 3:00pm to 5:00pm

Kelley Engineering Center, 1007
110 SW Park Terrace, Corvallis, OR 97331

Event Type

Lecture or Presentation

Event Topic

Research

Organization
College of Engineering, Electrical Engineering and Computer Science
Contact Name

Calvin Hughes

Contact Email

Calvin.Hughes@oregonstate.edu

Subscribe
Google Calendar iCal Outlook

Recent Activity