Transcript
Page 1: Internet Streaming Media Delivery:

1

Internet Streaming Media Delivery:

Zhen Xiao

Joint work with Lei Guo, Enhua Tan, Songqing Chen, Oliver Spatchcheck, and Xiaodong Zhang

Delving into

A quality and Resource Utilization Perspective

ACM SIGCOMM Internet Measurement Conference (IMC'06), October 2006

Page 2: Internet Streaming Media Delivery:

2

Multimedia on the Internet

• Education and research

• News media

• Entertainment and gaming

• Advertisement

Page 3: Internet Streaming Media Delivery:

3

Streaming Media

CDN/MDN

Streaming server

Page 4: Internet Streaming Media Delivery:

4

Pseudo Streaming

HTTP

http://www.YouTube.com/http://video.google.com/

meta file

Web server

Page 5: Internet Streaming Media Delivery:

5

Streaming Media

• Thousands of concurrent streams

• Flexible response to network congestion

• Efficient bandwidth utilization

• High quality to end users

• Challenges and techniques

Page 6: Internet Streaming Media Delivery:

6

Existing Measurements

• Access pattern and user behaviors– A bunch of measurement studies– Server clusters, media proxies

• Streaming mechanism and delivery quality– Few studies

• Traffic volume …– Downloading > pseudo streaming > streaming

(WWW’05, cookie talk 2005)– P2P >> all other media delivery systems

Page 7: Internet Streaming Media Delivery:

7

Our Measurement

• Investigate modern streaming services– The delivery quality and resource utilization

• Collect a large streaming media workload – From thousands of home users and business users– Hosted by a large ISP– Packet level instead of server logs

• Analyze commonly used streaming techniques– Automatic protocol switch– Fast Streaming– MBR encoding and rate adaptation

Page 8: Internet Streaming Media Delivery:

8

Outline

• Traffic overview

• Protocol rollover

• Fast Streaming

• Rate adaptation

• Conclusion

Page 9: Internet Streaming Media Delivery:

9

Traffic Overview

• User communities– Home user– Business user

• Media hosting services– Self-hosting– Third-party hosting

Page 10: Internet Streaming Media Delivery:

10

Number of requests

0.00%

10.00%

20.00%

30.00%

40.00%

50.00%

60.00%

70.00%

Home user Busi ness user

audi ovi deo

Business users access more audio than home users

Page 11: Internet Streaming Media Delivery:

11

On-demand media: File length

Business users tend to access longer audio/video files

Audio Video

pop songsmusic previews

Page 12: Internet Streaming Media Delivery:

12

On-demand media: Playback duration

Business users tend to play audio/video longer

Audio Video

pop songs

music previews

Page 13: Internet Streaming Media Delivery:

13

Live media: Playback duration

Business users tend to access live audio/video longer

Audio Video

Page 14: Internet Streaming Media Delivery:

14

Traffic Overview

• User communities– Home user– Business user– Working environment affects access pattern

• Media hosting services– Self-hosting– Third-party hosting

News and entertainment sites

Page 15: Internet Streaming Media Delivery:

15

Traffic Overview

• User communities– Business users tend to access streaming

media longer than home users– Working environment affects access pattern

• Media hosting services– Self-hosting– Third-party hosting

Page 16: Internet Streaming Media Delivery:

16

Media hosting services

Page 17: Internet Streaming Media Delivery:

17

Outline

• Traffic overview

• Protocol rollover

• Fast Streaming

• Rate adaptation

• Conclusion

Page 18: Internet Streaming Media Delivery:

18

Protocol RolloverStreaming server

RTSP/UDP

RTSP/TCP

HTTP/TCP Embed RTSP commands in HTTP packets

Media player

Traffic volume:

UDP: 23%

TCP: 77%

HTTP: rare

Page 19: Internet Streaming Media Delivery:

19

Protocol rollover time

Windows media service RealNetworks media service

Protocol rollover increases user startup time significantly

Startup latency = protocol rollover time + transport setup time + startup buffering time

Page 20: Internet Streaming Media Delivery:

20

Protocol selection and rollover avoidance

• Most streaming traffic are TCP-based– The usage of NAT?– MMS clients report private IP address in clear text

• Home user: 98.3% report 192.168.*.*• Business user: 89.5% report 192.168.*.*

• Protocol rollover sessions are minor– Home user: 7.37%– Business user: 7.95%

• Most streaming sessions use TCP directly– Why?

Page 21: Internet Streaming Media Delivery:

21

Protocol selection and rollover avoidance

• Windows media service– Specify the protocol in the media meta file

Use URL modifiers to avoid protocol rollover

Ex: rtspt://xxx.xxx.com:/xxx.wmv– More than 70%

• RealNetworks media service– NAT transversal techniques

Page 22: Internet Streaming Media Delivery:

22

Outline

• Traffic overview

• Protocol rollover

• Fast Streaming

• Rate adaptation

• Conclusion

Page 23: Internet Streaming Media Delivery:

23

Fast Streaming

• Fast Streaming: deliver media data “faster” than its encoding rate– Fast start– Fast cache– Fast recovery– Fast reconnect

• Always TCP-based

Page 24: Internet Streaming Media Delivery:

24

Media objects delivered with Fast Cache(VoD home user workload)

File length Encoding rate

Fast Cache is more widely used for media files with longer length and higher encoding rate.

Page 25: Internet Streaming Media Delivery:

25

Bandwidth Utilization

PLAY RTSP/1.0Bandwidth: 1.12 MbpsSpeed: 20.5

RTSP /1.0 200 OKSpeed: 5

Fast Cache Normal TCP

Page 26: Internet Streaming Media Delivery:

26

Fast Cache smooth bandwidth fluctuation

Rebuffer ratio = rebuffer time / play time

Fast Cache

Normal TCP

Page 27: Internet Streaming Media Delivery:

27

Fast Cache produces extra traffic

Most streaming sessions only request the initial part of a media object

Over supplied dataFast Cache: 55%

Normal TCP: 5%

Page 28: Internet Streaming Media Delivery:

28

Server response time

Third party media service Self-hosting media service

DESCRIBE foo.wmv RTSP/1.0

RTSP /1.0 200 OKSDP

RTT

SRT

sniffer

Page 29: Internet Streaming Media Delivery:

29

Server Load

Windows media load simulator

Windows Server 2003 Win XP

Server log

Ethernet

1 X 4 XSome CDNs/MDNs do not support Fast Cache at all

Page 30: Internet Streaming Media Delivery:

30

Outline

• Traffic overview

• Protocol rollover

• Fast Streaming

• Rate adaptation

• Conclusion

Page 31: Internet Streaming Media Delivery:

31

Rate Adaptation

96Kbps128Kbps320Kbps

… 1.128Mbps

Multiple-bit-rate encoding

Stream switch

WM: Intelligent streaming RM: SureStream

Stream thinning: deliver key frame only

Video cancellation

Page 32: Internet Streaming Media Delivery:

32

MBR encoding

on-demand audio

live audio

audio stream in video objects

video stream in video objects

42% on-demand video are MBR encoded

Page 33: Internet Streaming Media Delivery:

33

Stream switch

30 sec

60%

Streaming switch latency Low quality duration

3 sec

40%

Play-out buffer

Stream switch is often not smooth

Page 34: Internet Streaming Media Delivery:

34

Stream thinning

30 sec

70%

Page 35: Internet Streaming Media Delivery:

35

Fast Cache and stream switch

Do not work with each other: fewer stream switches than MBR encoded objects

playing buffering playing buffering bufferingplaying

5 sec

When network congestion occurs …

Like pseudo streaming

When rebuffer occurs

time

fill play-out buffer

Page 36: Internet Streaming Media Delivery:

36

Streaming quality and playback duration

Home user business user

Longer duration sessions have higher prob. of quality degradation

Business user workload has more quality degradation

>100 sec

88%

Page 37: Internet Streaming Media Delivery:

37

Streaming quality summary

The quality of media streaming on the Internet leaves much to be improved

Page 38: Internet Streaming Media Delivery:

38

Coordinating caching and rate adaptation

• Fast Cache: aggressively buffer data in advance– Over-utilize CPU and bandwidth resources– Neither performance effective nor cost-efficient

• Rate adaptation: conservatively switch to lower bit rate stream– Switch handoff latency

• Coordinated Streaming

Upper boundPrevent aggressive buffering

Lower boundPrevent switch latency

Page 39: Internet Streaming Media Delivery:

39

Coordinated Streaming

Rebuffering ratio Over-supplied data Switch latency

Page 40: Internet Streaming Media Delivery:

40

Conclusion

• Quality of Internet streaming– Often unsatisfactory– Need to improve

• Modern streaming media services– Over-utilize CPU and bandwidth resources– Not a desirable way to improve quality

• Coordinated Streaming– Combine merits of both caching and rate adaptation– Simple but effective


Recommended