Authors: Xinyan Zhang', Jiangchuan Liu, Bo Li, and TakShing Peter Yum'

1
CoolStreaming/DONet A Data-driven Overlay
Network for Peer-to-Peer Live Media Streaming

• Authors Xinyan Zhang., Jiangchuan Liu, Bo Li,
  and Tak-Shing Peter Yum.
• Presenter Elaine

2
Outline

• Introduction
• Related Work
• Design and Optimization of Donet
• Planet-Based Performance Evaluation
• CoolStreaming
• Conclusion and Future works

3

• Multimedia applications
• Live media streaming from a source to a large
  population of users.
• NetTV, news broadcast
• IP multicast is the best way
• The lack of incentives to install multicast
  routers
• Political and practical issues
• Application-level solution Overlay network
• Built on the unicast tunnels across cooperative
  participating users.

4

• Features of overlay network with streaming
  application.
• Easily crash/leave
• Streaming application
• High bandwidth
• Continuity
• Many overlay construction algorithms advocate a
  tree structure for data delivering.
• highly vulnerable.
• Restriction of the flow directions
• Proposed data-centric overlay network
• No predetermined roles
• Self-organization
• Data availability guides the flow directions.

5
Outline

• Introduction
• Related Work
• Design and Optimization of Donet
• Planet-Based Performance Evaluation
• CoolStreaming
• Conclusion and Future works

6
A brief overview of the existing overlay
streaming protocols in the pure p2p paradigm

• Tree-based protocol
• An internal node with high load leaving or
  crashing often causes underflow in a large
  population of descendants.
• Unbalanced load
• Repairing mechanisms are complex.

7

• Gossip-based protocol
• A node sends a newly generated message to a set
  of randomly selected nodes, iteratively.
• Cause redundancy especially for media streaming
  application.
• Partially used for membership management and
  data delivery in Donet

8
Outline

• Introduction
• Related Work
• Design and Optimization of Donet
• Planet-Based Performance Evaluation
• CoolStreaming
• Conclusion and Future works

9

• Three key modules
• Membership manager
• Partnership manager
• Scheduler

10

• Interaction of modules
• Node join and Membership Management
• Each Donet node has a unique id and Mcache.
• How to create and update the mcache.
• Each node periodically generates a membership
  message to announce its existence
• SCAM(Gossip) to distribute the mcache messages..
• Message format ltseq num, id,num partner, time to
  livegt
• Entry Message format ltseq num, id,num partner,
  time to live,last_update_timegt
• Node join algorithm

11

• Buffer Map Representation and Exchange
• A video stream is divided into segments of
  uniform length
• Buffer Map represent the availability of the
  segments
• Continuously exchange BM with partners, and
  schedules which segment is to be fetched from
  which partner accordingly.

12
(No Transcript)
13

• Failure Recovery and Partnership Refinement
• Failure Recovery
• Graceful departure
• Node failure
• Partnership Refinement
• Periodically establish new partnerships with
  nodes randomly selected from its mCache.
• Node i use function maxsi,j , sj,i to score
  its partners
• si,j the average number of segments that
  node i retrieved from node j per unit time.
• sj,i the average number of segments that node
  j retrieved from node j per unit time.

14
Outline

• Introduction
• Related Work
• Design and Optimization of Donet
• Planet-Based Performance Evaluation
• CoolStreaming
• Conclusion and Future works

15

• Design of the Experiment System
• Almost all active nodes over PlanetLab,200300
• Experiment May to June, 2004
• Each active PlanetLab node runs a copy of the
  prototyped program,acting a s a DONet node.
• Origin node Located in the United States
• Monitoring node In Hong Kong

16

• Performance under Stable Environment
• All nodes persist in the lifetime of the
  streaming(120min)
• Control overhead(left)
• Scalability(left)
• Playback continuity(right).

17

• Performance under Dynamic Environment
• Dynamic node joining , leaving, and failure.
• Following the ON/OFF model.

18

• Comparison with the Tree-based Overlay
• Metrics End-to-End delay / Continuity
• The number of partners is set to 4 in Donet.
• The degree of each tree node is limited to 3,
  except root(4).
• Different capacity and bandwidth cause some
  children move to lower levels.
• Fig 11/fig12/fig 13

19

• A full and balanced 3-ary tree of 231 nodes has
  a height of 5 only.

20
(No Transcript)
21

• Summary and Caveats
• Performance of Donet is acceptable
• Overhead is low(1 of the video traffic)
• Scalability
• Compared to tree-based overly
• Continuity is much better under highly dynamic
  environment.
• End-to-end delay is comparable.
• Caveats
• Scalability
• Representability

22
Outline

• Introduction
• Related Work
• Design and Optimization of Donet
• Planet-Based Performance Evaluation
• CoolStreaming
• Conclusion and Future works

23

• A public Internet based DONet package, called
  CoolStreaming.
• Released the first version (v.0.9) on May 30,
  2004. CoolStreaming v.0.9

24

• Two interesting facts
• The current Internet has enough available
  bandwidth to support TV-quality streaming(gt450
  Kbps)
• The larger the data-driven overlay is, the better
  the streaming quality it delivers.

25
Conclusion

• Proposed DONet
• Simple overlay structure (according to the data
  availability)
• Scalable membership and partnership management
  algorithm
• Good playback quality under formidable network
  condition
• Compared to tree-based overlay,it achieves much
  more continuous streaming with comparable delay.
• Released CoolStreaming v.0.9