1 / 8 network characteristics of video streaming traffic ashwin rao, yeon-sup lim, chadi barakat,...

15
1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project Planete Sophia Antipolis Univ. of Massachusetts Amherst

Post on 19-Dec-2015

215 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

1 / 8

Network Characteristics of Video Streaming Traffic

Ashwin Rao, Yeon-sup Lim, Chadi Barakat,

Arnaud Legout, Don Towsley, and Walid Dabbous

INRIA Project PlaneteSophia Antipolis

Univ. of Massachusetts 

Amherst

Page 2: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

2 / 8

Video Streaming in the Internet

20% to 40 % of Internet Traffic

Adobe Flash

HTML5 Microsoft SilverlightContainers

Desktop Browsers Mobile Applications

Internet Explorer

Firefox Chrome iOS (native)

Android (native)

What are the Network Characteristics of Video Streaming Traffic ?

Page 3: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

3 / 8

Generic Behavior of Video StreamingD

ownl

oad

Am

ount

Time

Buf

ferin

gBlock Size

On

Off

Steady State

Average rate

during steady state

Page 4: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

4 / 8

Streaming Strategies Identified

No On Off Cycles

Long On Off Cycles Short On Off Cycles

OFF OFF

Page 5: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

5 / 8

Streaming Strategies Used

Service YouTube Netflix

Container Flash HTML5 Silverlight

IE 9 Short Short Short

Firefox Short No Short

Chrome Short Long Short

iOS (native) - Based on encoding rate

Short

Android (native)

- Long Long

Streaming strategy depends on typeof application and container

Page 6: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

6 / 8

Streaming Strategies Comparison

No On Off Long On Off Short On Off

Engineering

Complexity

No Complex

Engineering

Explicit support on server or client side

TCP Friendly Yes – TCP File Transfer

Yes – periodic File Transfer

Unknown traffic not ack-

clocked

Receive buffer occupancy

Large Moderate Small

Unused bytes on user interruptions

Large amount

Moderate amount

Small amount

StrategyMetric

Page 7: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

7 / 8

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?

Page 8: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

8 / 8

THANK YOU

[email protected]

[email protected]

Page 9: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

9 / 8

BACKUPS

Page 10: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

10 / 8

YouTube Streaming StrategiesContainer Flash HTML5

Application Any Web

Browser

IE 9 Firefox Google Chrome

iOS (native)

Android (native)

Strategy Short Short No Long Depends 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

Page 11: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

11 / 8

Open Questions for 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?

[email protected]@inria.fr

Page 12: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

12 / 8

Netflix Streaming StrategiesContainer Silverlight 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 3 MB 4.5 to 6 MB

Page 13: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

13 / 8

Data Set and Locations

YouTube NetFlix

Desktop Mobile Desktop Mobile

5000 Flash,

3000 HTML5,

2000 HD

50 200 50

2 Sites in France and

2 Sites in USA

2 Sites in USA

Page 14: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

14 / 8

Conclusions

• Streaming Strategy depends on browser and container• Migration from one strategy to another can have a non-

negligible impact

Raw File Transfer

vs

Periodic Buffering

vs

No ack-clock

Page 15: 1 / 8 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim, Chadi Barakat, Arnaud Legout, Don Towsley, and Walid Dabbous INRIA Project

15 / 8