1 / 18 network characteristics of video streaming traffic ashwin rao, yeon-sup lim *, chadi barakat,...

19
1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao , Yeon-sup Lim * , Chadi Barakat , Arnaud Legout , Don Towsley * , and Walid Dabbous INRIA Sophia Antipolis, France * University of Massachusetts Amherst, USA

Upload: avery-dillon

Post on 27-Mar-2015

217 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

1 / 18

Network Characteristics of Video Streaming Traffic

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

†INRIA Sophia Antipolis, France

*University of Massachusetts Amherst, USA

Page 2: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

2 / 18

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

Page 3: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

3 / 18

Video Streaming Services

Containers

Desktop Browsers Native Mobile Applications

What are the Network Characteristics of Video Streaming Traffic?

Page 4: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

4 / 18

Objective• Network Characteristics of Video Streaming Traffic

– Causes– Impact

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

Page 5: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

5 / 18

Datasets

• YouTube videos– 5000 Flash– 3000 HTML5– 2000 HD Videos (Flash)– 50 Mobile

• Netflix videos - Silverlight– 200 to Desktop – 50 to Mobile

Page 6: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

6 / 18

Measurement Technique

802.11

Packet Capture

Page 7: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

7 / 18

Measurement Locations

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

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

YouTube

YouTube and

Netflix

Similar Traffic Characteristics at Each Location

Page 8: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

8 / 18

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

Page 9: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

9 / 18

Generic Behavior of Video StreamingD

ownl

oad

Amou

nt

Time

Buf

ferin

gBlock Size

On

Off

Steady State

Average rate

Video encoding rate

Page 10: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

10 / 18

We Identified Three Streaming Strategies

No On Off Cycles

Long On Off Cycles

OFF OFF

Short On Off Cycles

Streaming strategies vastly different

Page 11: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

11 / 18

Streaming Strategies UsedService 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 withapplication type and container

Page 12: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

12 / 18

Important Features of a Strategy

• Buffering Amount• Block Size• Accumulation Ratio

Average download rate in steady state phase

Video encoding rate=

Page 13: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

13 / 18

Short ON OFF Strategy

64 kB

40 sec. of playback

Server side rate control with

absence of ACK clocks

1.25Buffering

independent ofencoding rate

Browser throttles rate

256 kB

Significant differences between implementations

Page 14: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

14 / 18

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

Page 15: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

15 / 18

Impact of Streaming StrategiesNo 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

StrategyMetric

Page 16: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

16 / 18

Model for Aggregate Rate of Streaming Traffic

• Objective– Capture statistical properties of aggregate

streaming trafficBarakat et al., A flow-based model for Internet backbone traffic, In IMW’02.

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

Page 17: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

17 / 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

Page 18: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

18 / 18

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

Page 19: 1 / 18 Network Characteristics of Video Streaming Traffic Ashwin Rao, Yeon-sup Lim *, Chadi Barakat, Arnaud Legout, Don Towsley *, and Walid Dabbous INRIA

THANK YOU

Network Characteristics of Video Streaming Traffic