Network Characteristics of Video Streaming Traffic

1
Network Characteristics of Video Streaming Traffic

• Ashwin Rao, Yeon-sup Lim, Chadi Barakat,
• Arnaud Legout, Don Towsley, and Walid Dabbous

University of Massachusetts  Amherst, USA
INRIA Sophia Antipolis France
2
Video Streaming Services


Containers
What are the Network Characteristics of Video
Streaming Traffic?
Desktop Browsers Desktop Browsers Desktop Browsers Native Mobile Applications Native Mobile Applications

3
Objective

• What exactly happens during video streaming?
• Arrival of data packets
• Strategies to stream videos
• Potential Impact

4
Outline

• Introduction and Motivation
• Datasets and Measurement Techniques
• Streaming Strategies
• Impact of Streaming Strategies

5
Datasets

• YouTube videos
• Flash, HTML5, and HD (Flash)
• Mobile
• Netflix videos - Silverlight
• Desktop
• Mobile

6
Measurement Technique
Packet Capture
802.11
7
Measurement Locations

• France
• Academic (Wired Wi-fi for mobile)
• Residential (Wi-fi)
• USA
• Academic (Wired Wi-fi for mobile)
• Residential (Wired)

YouTube
Similar Traffic Characteristics at Each Location
YouTube and Netflix
8
Outline

• Introduction and Motivation
• Datasets and Measurement Techniques
• Streaming Strategies
• Impact of Streaming Strategies

9
Generic Behavior of Video Streaming
Steady State
Off
On
Average rate ? Video encoding rate
Block Size
Download Amount
Buffering
Time
10
We Identified Three Streaming Strategies
No On Off Cycles
Streaming strategies vastly different
Short On Off Cycles
Long On Off Cycles
OFF
OFF
11
Streaming Strategies Used
Service YouTube YouTube YouTube Netflix
Container Flash HD (Flash) HTML5 Silverlight
IE 9 Short No Short Short
Firefox Short No No Short
Chrome Short No Long Short
iOS (native) - - Based on encoding rate Short
Android (native) - - Long Long
Streaming strategy differs with application type
and container
12
Features Controlling Arrival of Data Packets

• Buffering Amount
• Block Size
• Accumulation Ratio

Average download rate in steady state phase

Video encoding rate
13
Arrival of Packets for Short ON OFF Strategy
1.25
64 kB
Buffering independent of encoding rate
40 sec. of playback
Significant differences between implementations
Browser throttles rate
Server side rate control with absence of ACK
clocks
256 kB
14
Outline

• Introduction and Motivation
• Datasets and Measurement Techniques
• Streaming Strategies
• Impact of Streaming Strategies

15
Impact of Streaming Strategies
No On Off Long On Off Short On Off
TCP Friendly Yes TCP File Transfer Yes Periodic File Transfer Unknown traffic not ack-clocked
Playout buffer occupancy Large Moderate Small
Unused bytes on user interruptions Large amount Moderate amount Small amount
Strategy
Metric
16
Model for Aggregate Rate of Streaming Traffic

• Objective
• Capture statistical properties of aggregate
  streaming traffic

Barakat et al., A flow-based model for Internet
backbone traffic, In IMW02.

• Uses
• Dimension the network
• Quantify impact of user interruptions

17
Aggregate Rate of Video Streaming Traffic
Aggregate Rate
Amount of data downloaded
Arrival Rate of streaming sessions (Poisson)
18
Insights from Model

• No User Interruptions
• Aggregate rate (mean, variance, etc.) independent
  of streaming strategy
• Dimensioning rules do not change
• Strategy to optimize other goals (server load,
  etc.)
• Users Interruptions
• Impact of buffering amount and accumulation ratio
  on wasted bandwidth

19
Summary

• Most popular clients and containers for video
  streaming
• Streaming strategy differs with client
  applications and container
• HTML5 streaming vastly differs with client
  applications
• Model to study impact of streaming strategies

20
Open Questions for the CCN community

• Should CCN nodes be aware of the underlying
  streaming strategy?
• What is the optimal streaming strategy for CCN?
• Is there an optimal caching strategy for a given
  streaming strategy?
• What is the impact of user interruptions due to
  lack of interest on CCN caches?

21
THANK YOU
Network Characteristics of Video Streaming Traffic

• ashwin.rao_at_inria.fr

22
BACKUPS
23
Short or Long

• Block Size Threshold 2.5 MB

Long
Short
Long
OFF
24
ACK Clocks

• Source sends packets on receiving ACK
• ACKs as an indication of available bandwidth

46 packets sent in the first RTT after an OFF
period of more than 500 ms
25
Conclusion

• Most popular clients and containers for video
  streaming
• Streaming strategy differs with client
  applications and container
• HTML5 streaming vastly differs with client
  applications
• Model to study impact of streaming strategies

26
User Interruptions
Video download will be in progress when
Playback time downloaded in buffering phase
Video duration
gt
1
Accumulation ratio
Fraction of video watched
-
X
27
Impact of Losses

• Merging of cycles
• Playback can freeze
• Longer buffering phase

28
HTML5

• Primary - webM
• Very few - h.264

29
Netflix Streaming Strategies
Container Silverlight Silverlight for Mobile Devices Silverlight for Mobile Devices
Application Any Web Browser iOS (native) Android (native)
Strategy Short Short Long
Buffering Amount 30 MB to 150 MB 10 to 20 MB 35 to 45 MB
Block Size 0.5 MB to 2 MB 0.5 to 2.5 MB 4.5 to 6 MB
30
YouTube Streaming Strategies
Container Flash HTML5 HTML5 HTML5 HTML5 HTML5
Application Any Web Browser IE 9 Firefox Google Chrome iOS (native) Android (native)
Strategy Short Short No Long Multiple Long
Buffering Amount 40 s Up to 15 MB Video Size Up to 15 MB 40 s of playback or up to 20 MB Up to 10 MB
Block Size 64 kB 256 kB NA 5 MB to 8 MB 64 kB 2 MB to 8 MB
31
Tradeoff

• Migration from one strategy to another can have a
  non-negligible impact
• Raw File Transfer
• vs
• Periodic Buffering
• vs
• No ack-clock

32
Video Streaming in the Internet

• 20 to 40 of all Internet traffic
• Traffic share steadily increasing in recent years
• Streaming over HTTP using TCP
• Firewall configurations
• TCP flows assumed to be fair