Measurements of the Congestion Responsiveness of Windows Streaming Media

1
Measurements of the Congestion Responsiveness of
Windows Streaming Media
James Nichols, Mark Claypool, Robert Kinicki and
Mingzhe Li Computer Science Department Worcester
Polytechnic Institute Massachusetts, USA
NOSSDAV 2004 - Kinsale, County Cork, Ireland
2
Network Impact of Streaming Media

• Streaming Media has specific bitrate and timing
  requirements.
• Flows unresponsive to congestion can cause
  unfairness and contribute to congestion collapse.
• Streaming Media applications rely on the
  application layer to provide adaptability to
  available capacity
• Windows Streaming Media (WSM) uses Intelligent
  Streaming

NOSSDAV 2004
3
Methodology

• Construct testbed
• Create/adapt tools
• Encode content
• Systematic control

• Examine SBR clip
• Range of SBR clips
• MBR clips
• Vary loss and latency

NOSSDAV 2004
4
Single Bitrate Clip Experiment

• Hypothesis SBR clips are unresponsive to
  congestion
• Latency 45 ms
• Induced loss 0
• Bottleneck capacity 725 Kbps
• Start a TCP flow through the link
• 10 Seconds later stream a WSM clip
• Measure achieved bitrates and loss rates for each
  flow

NOSSDAV 2004
5
340 Kbps Clip - Bottleneck Capacity 725 Kbps
TCP- Friendly?
lt 0.001 packet loss After 15 seconds
6
548 Kbps Clip - Bottleneck Capacity 725 Kbps
Not TCP- Friendly!
0.003 packet loss for WSM 0.006 packet loss
for TCP after 15 seconds
7
1128 Kbps Clip - Bottleneck Capacity 725 Kbps
Responsive To Capacity?
8
SBR Clips - Bottleneck Capacity 725
KbpsBuffering Period
NOSSDAV 2004
9
SBR Clips - Bottleneck Capacity 725 KbpsPlayout
Period
NOSSDAV 2004
10
Multiple Bitrate Clips

• Hypothesis Multiple Bitrates make WSM more
  responsive to congestion
• Same experiment as before, but with different
  encoded content
• Vary network capacity 250, 725, 1500 Kbps
• Created two sets of 10 multiple bitrate clips
• Experiments with lots of other MBR clips

NOSSDAV 2004
11
Multiple Bitrate Content

• First set of clips (adding lower)
• 1128 Kbps
• 1128-764 Kbps
• 1128-764-548 Kbps
• 1128-764-548-340-282-148-109-58-43-28 Kbps

NOSSDAV 2004
12
Multiple Bitrate Clips
Buffering
Playout
NOSSDAV 2004
13
Additional experiments

• Not enough time to discuss all the results
• Different bottleneck capacities
• Vary loss rate and latencies
• Also looked at other network level metrics
  interarrival times, and burst lengths.

NOSSDAV 2004
14
Conclusions

• WSM is not CBR
• Buffering period
• Not TCP-friendly
• Responds to capacity
• Chooses highest rate less than capacity
• Otherwise, lowest rate available
• Performance depends upon encoding

NOSSDAV 2004
15
Future Work

• Run the same experiments with other streaming
  technologies RealVideo and Quicktime
• Examine the effects of different content types
• Build NS simulation model of streaming media for
  use in future research

NOSSDAV 2004
16
Measurements of the Congestion Responsiveness of
Windows Streaming Media
James Nichols, Mark Claypool, Robert Kinicki and
Mingzhe Li http//perform.wpi.edu
NOSSDAV 2004 - Kinsale, County Cork, Ireland