ppt

18
1 PV'2003, Nantes France, April 2003 Measurement of the Congestion Responsiveness of RealPlayer Streaming Video Over UDP Jae Chung, Mark Claypool, Yali Zhu Computer Science Department Worcester Polytechnic Institute Worcester, MA, USA http://www.cs.wpi.edu/~claypool/papers/h2h/

Upload: ronny72

Post on 14-Jun-2015

165 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: PPT

1PV'2003, Nantes France, April 2003

Measurement of the Congestion Responsiveness

of RealPlayer Streaming Video Over UDP

Jae Chung, Mark Claypool, Yali Zhu

Computer Science DepartmentWorcester Polytechnic Institute

Worcester, MA, USA

http://www.cs.wpi.edu/~claypool/papers/h2h/

Page 2: PPT

PV'2003, Nantes France, April 20032

The Hungry Wolf – Bandwidth Requirements for

Video•Voice uses 32-64 Kbps, but video has

broader range of data rates– Videoconference 0.1 Mbps (H.261, MPEG-4)– VCR 1.2 Mbps (MPEG-1)– Broadcast quality 2-4 Mbps (MPEG-2)– HDTV quality 25-34 Mbps (MPEG-2)

Thus, potential for more than network capacity

Page 3: PPT

PV'2003, Nantes France, April 20033

The Wolf on the Prowl – Transport Protocols for Video

•Streaming video doesn’t always like TCP– Wants rate-based not window-based– Can tolerate some loss– AIMD causes rate fluctuationsSo, uses UDP where application controls

•But UDP has no built-in congestion control– Unfair, unfriendly, and even collapse!

•Approaches to have router detect and limit– Often model video as CBR “firehose” (is it?)

Page 4: PPT

PV'2003, Nantes France, April 20034

The Wolf Pack – Commercial Video

•Commercial products have major impact

•Have been studies characterizing commercial traffic– Bandwidth use, frame rate, user use…

•But no work measuring responsiveness, or lack of it, of commercial video products

Measure responsiveness of RealPlayer over UDP

Page 5: PPT

PV'2003, Nantes France, April 20035

Outline

•Introduction •Background •Experiments

•Analysis

•Conclusions

Page 6: PPT

PV'2003, Nantes France, April 20036

RealVideo Network Characteristics

•SureStream (is it effective?)– Allows bandwidth scaling

•Buffer (how fast is it filled?)– Smoother playout

Server

RTSP

Data: TCP or UDP

-Choice automatic

-We force

Page 7: PPT

PV'2003, Nantes France, April 20037

Outline

•Introduction •Background

•Experiments •Analysis

•Conclusions

Page 8: PPT

PV'2003, Nantes France, April 20038

Responsiveness Measurement Environment

• Encoding rates RealTracker

• Data rates via tcpdump

• Loss rates and round-trip times via ping

• TBF to limit bandwidth– 75, 150, 300, 600 kbps

• 79 (of 100) Clips, 2 measurements per clip

Internet

Client

Hub

Router

10Mbps

DSL

700KbpsClient

Token Bucket Filter

UDP

TCPServer

Page 9: PPT

PV'2003, Nantes France, April 20039

Outline

•Introduction •Background

•Experiments

•Analysis

– Bandwidth – TCP-Friendly– Media Scaling– Buffering Rate

•Conclusions

Page 10: PPT

Head-to-Head Bandwidth

0

50

100

150

200

250

300

350

400

0 50 100 150 200 250 300 350 400

Ave

rag

e B

and

wid

th fo

r U

DP

(K

bps

)

Average Bandwidth for TCP (Kbps)

75 Kbps150 Kbps300 Kbps600 Kbps

Page 11: PPT

PV'2003, Nantes France, April 200311

A Sheep in Wolf’s Clothing?

•Remove low bandwidth (36%)

•Then remove unscalable (14%)

[PFTK98]

Page 12: PPT

Media Scaling Distribution

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 6 7 8 9

Cum

ula

tive

Den

sity

Number of Scales (Coded-Bandwidth)

Page 13: PPT

Media Scaling Dynamics

0

50000

100000

150000

200000

250000

0 50 100 150 200 250 300 350

Co

ded-

Ban

dw

idth

(bp

s)

P layout + Buffereing Tim e (sec)

local bw lim it (35 kbps)Scale (Coded-BW ) Movement for C lip-65 TCPScale (Coded-BW ) Movement for C lip-65 UDP

Page 14: PPT

Media Scaling Adaptation Speed

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10 20 30 40 50 60 70 80

Cum

ula

tive

Den

sity

E laps Time in Seconds: 0 to tim e(Coded-BW < 35kbps)

Scale Adaptation Speed for TCPScale Adaptation Speed for UDP

Page 15: PPT

Buffering Rate to Playout Rate

0

1

2

3

4

5

6

7

0 50 100 150 200 250 300 350 400

Ave

rage

Buf

feri

ng R

ate

/ A

vera

ge S

tead

y P

layo

ut R

ate

Average Steady Playout Rate (Kbps)

TCP (A ll DSL-TBF Runs)UDP (All DSL-TBF R uns)

Page 16: PPT

PV'2003, Nantes France, April 200316

Summary

•No concrete incentives to respond to congestion– In fact, may be “rewarded” for not

•However, RealVideo clearly adjusts– Often TCP-Friendly

•Content providers need to provide chance for scalability– Scaling levels

•Buffering at higher rate is bad for network since bursty

•Why not TCP? API is limiting

Page 17: PPT

PV'2003, Nantes France, April 200317

Future Work

•Other commercial Players– Microsoft Media Player

•Perceptual quality of video over TCP versus UDP

•Characterization of clips on Internet– So can examine “typical” clips

Page 18: PPT

18PV'2003, Nantes France, April 2003

Measurement of the Congestion Responsiveness

of RealPlayer Streaming Video Over UDP

Jae Chung, Mark Claypool, Yali Zhu

Computer Science DepartmentWorcester Polytechnic Institute

Worcester, MA, USA

http://www.cs.wpi.edu/~claypool/papers/h2h/