ABSTRACT
This paper investigates HTTP streaming traffic from an ISP perspective. As streaming traffic now represents nearly half of the residential Internet traffic, understanding its characteristics is important. We focus on two major video sharing sites, YouTube and DailyMotion.
We use ten packet traces from a residential ISP network, five for ADSL and five for FTTH customers, captured between 2008 and 2011. Covering a time span of four years allows us to identify changes in the service infrastructure of some providers.
From the packet traces, we infer for each streaming flow the video characteristics, such as duration and encoding rate, as well as TCP flow characteristics. Using additional information from the BGP routing tables allows us to identify the originating Autonomous System (AS). With this data, we can uncover: the server side distribution policy, the impact of the serving AS on the flow characteristics and the impact of the reception quality on user behavior.
A unique aspect of our work is how to measure the reception quality of the video and its impact on the viewing behavior. We see that not even half of the videos are fully downloaded. For short videos of 3 minutes or less, users stop downloading at any point, while for videos longer than 3 minutes, users either stop downloading early on or fully download the video. When the reception quality deteriorates, fewer videos are fully downloaded, and the decision to interrupt download is taken earlier.
We conclude that (i) the video sharing sites have a major control over the delivery of the video and its reception quality through DNS resolution and server side streaming policy, and (ii) that only half of the videos are fully downloaded and that this fraction dramatically drops when the video reception quality is bad.
- Vijay Kumar Adhikari, Sourabh Jain, and Zhi-Li Zhang. YouTube Traffic Dynamics and Its Interplay with a Tier-1 ISP: An ISP Perspective. In IMC, 2010.Google ScholarDigital Library
- Vijay Kumar Adhikari, Sourabh Jain, and Zhi Li Zhang. Where Do You "Tube"? Uncovering YouTube Server Selection Strategy. In IEEE ICCCN, 2011.Google Scholar
- Alexa. http://www.alexa.com/siteinfo/dailymotion.com.Google Scholar
- Meeyoung Cha, Haewoon Kwak, Pablo Rodriguez, Yong-Yeol Ahn, and Sue Moon. I Tube, You Tube, Everybody Tubes: Analyzing the World's Largest User Generated Content Video System. In IMC, 2007. Google ScholarDigital Library
- Xu Cheng, Cameron Dale, and Jiangchuan Liu. Understanding the Characteristics of Internet Short Video Sharing: YouTube as a Case Study. In IMC, 2007.Google Scholar
- comScore. http://www.comscore.com/press_events/press_releases/2009/3/youtube_surpasses_100_million_us_viewers.Google Scholar
- Florin Dobrian, Asad Awan, Ion Stoica, Vyas Sekar, Aditya Ganjam, Dilip Joseph, Jibin Zhan, and Hui Zhang. Understanding the Impact of Video Quality on User Engagement. In SIGCOMM, 2011. Google ScholarDigital Library
- Jeffrey Erman, Alexandre Gerber, Mohammad Taghi Hajiaghayi, Dan Pei, and Oliver Spatscheck. Network-aware forward caching. In WWW, 2009. Google ScholarDigital Library
- Alessandro Finamore, Marco Mellia, Maurizio Munafo, Ruben Torres, and Sanjay Rao. YouTube Everywhere: Impact of Device and Infrastructure Synergies on User Experience. In IMC, 2011. Google ScholarDigital Library
- Phillipa Gill, Martin Arlitt, Zongpeng Li, and Anirban Mahanti. Characterizing User Sessions on YouTube. In ACM/SPIE MMCN, 2008.Google Scholar
- Phillipa Gill, Martin F. Arlitt, Zongpeng Li, and Anirban Mahanti. Youtube Traffic Characterization: A View From the Edge. In IMC, 2007. Google ScholarDigital Library
- HTB. http://luxik.cdi.cz/~devik/qos/htb/.Google Scholar
- Parikshit Juluri, Louis Plissonneau, and Deep Medhi. Pytomo: A Tool for Analyzing Playback Quality of YouTube Videos. In ITC23, 2011. Google ScholarDigital Library
- Gregor Maier, Anja Feldmann, Vern Paxson, and Mark Allman. On Dominant Characteristics of Residential Broadband Internet Traffic. In IMC, 2009. Google ScholarDigital Library
- Marcin Pietrzyk, Jean-Laurent Costeux, Guillaume Urvoy-Keller, and Taoufik En-Najjary. Challenging Statistical Classification for Operational Usage: the ADSL Case. In IMC, 2009. Google ScholarDigital Library
- Marcin Pietrzyk, Louis Plissonneau, Guillaume Urvoy-Keller, and Taoufik En-Najjary. On Profiling Residential Customers. In TMA, 2011. Google ScholarDigital Library
- Louis Plissonneau, Taoufik En-Najjary, and Guillaume Urvoy-Keller. Revisiting Web Traffic from a DSL Provider Perspective: the Case of YouTube. In ITC SS on Network Usage and Traffic, 2008.Google Scholar
- Ingmar Poese, Benjamin Frank, Bernhard Ager, Georgios Smaragdakis, and Anja Feldmann. Improving Content Delivery using Provider-aided Distance Information. In IMC, 2010. Google ScholarDigital Library
- Pytomo. http://code.google.com/p/pytomo/.Google Scholar
- R. Nelson S. Alcock. Application Flow Control in YouTube Video Streams. ACM SIGCOMM CCR, 41(2), 2011. Google ScholarDigital Library
- Ruben Torres, Alessandro Finamore, Jesse Kim, Marco Mellia, Maurizio Munafo, and Sanjay Rao. Dissecting Video Server Selection Strategies in the YouTube CDN. In IEEE ICDCS, 2011. Google ScholarDigital Library
- Tstat. http://tstat.tlc.polito.it/.Google Scholar
Index Terms
- A longitudinal view of HTTP video streaming performance
Recommendations
Are we in the middle of a video streaming revolution?
Special Sections on the 20th Anniversary of ACM International Conference on Multimedia, Best Papers of ACM Multimedia 2012It has been roughly 20 years since the beginning of video streaming over the Internet. Until very recently, video streaming experiences left much to be desired. Over the last few years, this has significantly improved making monetization of streaming, ...
An encoding-aware bitrate adaptation mechanism for video streaming over HTTP
AbstractThe great interest in flix-like services has amplified multimedia traffic over the Internet. Recently released traffic forecasting predicts that video-related traffic will be responsible for the majority of Internet traffic by 2022. Such traffic ...
Araneola: A scalable reliable multicast system for dynamic environments
This paper presents Araneola (Araneola means ''little spider'' in Latin.), a scalable reliable application-level multicast system for highly dynamic wide-area environments. Araneola supports multi-point to multi-point reliable communication in a fully ...
Comments