1. Enabling Contribution Awareness in an Overlay Broadcasting System
ACM SIGCOMM 2006
Presented by He Yuan

2. Outline Background Related Work Contribution-aware Design
Implementation and Experiments
Conclusion
Discussion

3. Video Broadcast using Overlay Multicast
Encoder E
Boston D E
Pisa D
D: DSL
E: Ethernet D
San Francisco
Tokyo E LA
Overlay Tree
Boston
NYC
Pisa LA
San Francisco
Tokyo

4. Background I State-of-Art in Overlay Broadcast
Architecture and Protocol Design
Narada, SplitStream, CoopNet, DONet ...
Significant progress in scalability & resiliency
Real Deployments
ESM*, CoolStreaming, PPLive, SopCast ...

5. Background II Much success to date:
Homogeneous environments with abundant bandwidth
Heterogeneity in node upload bandwidth
Upload access bandwidth varies widely
Hosts may choose to forward differently
Insufficient bandwidth resource
Download
Upload
DSL
Kbps
64-256Kbps
Cable
1-6Mbps
Kbps
Ethernet
≥ 10Mbps
> 80%
< 20%

6. Related Work Bit-for-bit policy Differential Admission Control
Effective only in BT-like systems
Differential Admission Control
Not feasible in the mainstream Internet
Taxation model
Incentive vs. Fairness

7. Goals and Challenges Goals Challenges Good utilization of bandwidth
Differential and equitable distribution
Guaranteed QoS
Challenges
More generic than bit-for-bit policy
Distributed sampling and computing
Dynamic environment

8. Contribution-aware Design
Assumptions
Multi-tree-based data dissemination
Bandwidth distribution policy
System design

9. Assumptions Abundant download bandwidth
Different levels of contribution
Actual contribution fi reflected by Forwarding bound Fi
Non-strategic honest clients
To encourage a host to relax its Fi

10. Multi-tree-based data dissemination
Using MDC, split into T-equally sized stripes
T trees, each distributes a single stripe of size S/T
Overall quality depends on the number of stripes received
Number of trees node i is entitled to =
S Kbps
Peer A
Source
Peer C
S/3
S/3
S/3
Tree 1
Tree 2
Tree 3

11. Bandwidth distribution policy
∑ fj / N j
Entitled bandwidth
Contribution
0 < α < 1
More generic than bit-for-bit
Differential and Equitable Distribution

12. Bandwidth distribution: Example
S = 400Kbps T = 4 avg f = 300Kbps α = 0.5
fE = 500Kbps fD = 100Kbps
rE = 0.5* *300 = 400Kbps  entitled to 4 trees
rD = 0.5* *300 = 200Kbps  entitled to 2 trees
Entitled Node
Source
Excess Node
100Kbps
100Kbps
100Kbps
100Kbps E D D E E D E D

13. System Design Distributed System Sampling
Computing Number of Entitled Trees
Smoothing
Locating Excess Bandwidth
Backoff in Excess Tree
Contribution-Aware Node Prioritization

14. Implementation and Experiments
Use Slashdot to evaluate 2 systems:
Cont-Agnostic: multi-tree broadcast system
Cont-Aware: multi-tree + contribution-aware heuristics
S=400Kbps, T=4, stripe size S/T=100Kbps
2 types of peers: Ethernet fmax ≤800Kbps, DSL fmax ≤100Kbps
HC: Kbps, LC: Kbps
Broadcast Event
DSL (100Kbps)
Ethernet (10Mbps)
Peak Group Size
SIGCOMM2002
48%
52% 78
SOSP2003 54
Rally
75%
25%
481
Slashdot
73%
27%
158
GrandChallenge
82%
18%
276
Conferences
Mainstream Internet

15. Evaluation Goals
Fairness
Overall quality of playback
Stability

16. Performance: High Contributors
System
Mean
Std. Dev
Cont-Agnostic
353
60.9
Cont-Aware
415
24.6
Better
Cont-Aware gives HC better performance

17. Performance: Low Contributors
Better
System
Mean
Std. Dev
Cont-Agnostic
311
80.5
Cont-Aware
295
34.8
Better
Similar performance among similar contributors

18. Stability Time between Tree Reductions Reconnection time (in sec)
Cont-Aware performs slightly worse
Reductions => slight dips in quality
Not complete disconnection, 63.4% from 43, 34.1% from 32, only 2.5% from 21 and 10
Reconnection time (in sec)
Cont-Aware
Cont-Agnostic HC
7.1
80.82 LC
53.42
65.26
Overall
48.25
69.83

19. Performance across traces for high contributors

20. Conclusion Resource-scarce, heterogeneous environments
Two key ideas: Multi-trees and Linear Taxation
Provide fairness in overlay broadcasting in mainstream Internet environments

21. Discussion Applying MDC to Multi-tree overlay
The issue of redundancy in coding
What’s different in the resulting system?
More bandwidth resource or Better QoS
Incentive or fairness
Where to go?
Customized user requirement - Demand according to capacity
Location-aware streaming reuse technique

22. Thanks!