Claypool

Courses

Publications

Students

Projects

Service

Downloads

Misc


WPI (Worcester Polytechnic Institute)

Computer Science Department
------------------------------------------

Analyzing DASH Video Streaming over Geostationary Satellite Networks

Daniel Quackenbush, Kristi Prifti, and Zijian Guan

Advisor: Mark Claypool

Major Qualifying Project
Computer Science, WPI
Terms ABC 2022-2023

Abstract

Dynamic Adaptive Streaming over HTTP (DASH) is a popular, modern method for streaming media. Adaptive bitrate (ABR) functionality allows DASH to adjust to changing network conditions during the course of the stream. However, little work has assessed DASH's performance over geostationary satellite networks, where high latencies cause many network protocols to perform poorly. This paper presents results from experiments that evaluate DASH over a commercial geostationary satellite connection, comparing performance when adjusting two different configuration settings of the stream: The length of video segments, and the ABR algorithm used to make decisions. Results show that: 1) longer segment lengths are more stable but fail to adjust bitrate, staying at a low video quality; and 2) the buffer-based BOLA ABR algorithm has fewer stalls and better video quality than either a throughput-based approach or a hybrid approach.


Download


[WPI Homepage] [CS Homepage] [Mark Claypool's Homepage]