Your SlideShare is downloading. ×
0
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)
Upcoming SlideShare
Loading in...5
×

Thanks for flagging this SlideShare!

Oops! An error has occurred.

×
Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Applications (Part 3) - Peer-to-Peer Media Streaming (PPT)

1,127

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
1,127
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
36
Comments
0
Likes
0
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
No notes for slide

Transcript

  • 1. CS 414 – Multimedia Systems Design Lecture 40 – P2P Media Streaming William Conner Spring 2008 CS 414 - Spring 2008
  • 2. Outline <ul><li>Background </li></ul><ul><li>IP Multicast </li></ul><ul><li>Content delivery networks </li></ul><ul><li>P2P media streaming design </li></ul><ul><ul><li>Tree-based </li></ul></ul><ul><ul><li>Mesh-based </li></ul></ul><ul><li>Case study: PPLive </li></ul>CS 414 - Spring 2008
  • 3. Reading <ul><li>“ Opportunities and Challenges of Peer-to-Peer Internet Video Broadcast” by Liu et al. (Required) </li></ul><ul><li>“ Insights into PPLive: A Measurement Study of a Large-Scale P2P IPTV System” by Hei et al. (Recommended) </li></ul><ul><li>“ Mapping the PPLive Network: Studying the Impacts of Media Streaming on P2P Overlays” by Vu et al. (Recommended) </li></ul>
  • 4. Acknowledgements <ul><li>Some lecture material borrowed from the following sources </li></ul><ul><ul><li>Sanjay Rao’s lecture on P2P multicast in his ECE 695B course at Purdue </li></ul></ul><ul><ul><li>“ Insights into PPLive: A Measurement Study of a Large-Scale P2P IPTV System” by Hei et al. </li></ul></ul>
  • 5. Background <ul><li>Large-scale video broadcast over Internet </li></ul><ul><ul><li>Real-time video streaming </li></ul></ul><ul><ul><li>Need to support large numbers of viewers </li></ul></ul><ul><ul><ul><li>AOL Live 8 broadcast peaked at 175,000 (July 2005) </li></ul></ul></ul><ul><ul><ul><li>CBS NCAA broadcast peaked at 268,000 (March 2006) </li></ul></ul></ul><ul><ul><li>Very high data rate </li></ul></ul><ul><ul><ul><li>TV quality video encoded with MPEG-4 would require 1.5 Tbps aggregate capacity for 100 million viewers </li></ul></ul></ul><ul><ul><ul><li>NFL Superbowl 2007 had 93 million viewers in the U.S. (Nielsen Media Research) </li></ul></ul></ul>
  • 6. Possible Solutions <ul><li>Single server </li></ul><ul><li>IP multicast </li></ul><ul><li>Content delivery networks (CDNs) </li></ul><ul><li>Application end points (pure P2P) </li></ul>
  • 7. Single Server <ul><li>Application-layer solution </li></ul><ul><ul><li>Single media server unicasts to all clients </li></ul></ul><ul><li>Needs very high capacity to serve large number of clients </li></ul><ul><ul><li>CPU </li></ul></ul><ul><ul><li>Main memory </li></ul></ul><ul><ul><li>Bandwidth </li></ul></ul><ul><li>Impractical for millions of simultaneous viewers </li></ul>
  • 8. Single Server P C C C C C C C C C
  • 9. IP Multicast <ul><li>Network-layer solution </li></ul><ul><ul><li>Routers responsible for multicasting </li></ul></ul><ul><li>Efficient bandwidth usage </li></ul><ul><li>Requires per-group state in routers </li></ul><ul><ul><li>Scalability concern </li></ul></ul><ul><ul><li>Violates end-to-end design principle </li></ul></ul><ul><li>Slow deployment </li></ul><ul><ul><li>IP multicast is often disabled in routers </li></ul></ul><ul><li>Difficult to support higher layer functionality </li></ul>
  • 10. IP Multicast “ Smart Network” Berkeley Gatech Stanford Per-group Router State Source: Purdue Source: Sanjay Rao’s lecture from Purdue
  • 11. Overlay Network <ul><li>Consists of application-layer links </li></ul><ul><li>Application-layer link is logical link consisting of one or more links in underlying network </li></ul><ul><li>Used by both CDNs and pure P2P systems </li></ul>R1 R2 A B R3 C D
  • 12. Content Delivery Networks <ul><li>Strategically located replicas unicast content to nearby clients </li></ul><ul><ul><li>Reduces burden on primary server </li></ul></ul><ul><ul><li>Improves perceived performance at client </li></ul></ul><ul><li>Akamai CDN is the largest </li></ul><ul><ul><li>Reports peak aggregate capacity of 200 Gbps </li></ul></ul><ul><ul><li>Not enough for 1.5 Tbps requirement for 100 million simultaneous viewers </li></ul></ul><ul><li>Limelight CDN served YouTube content </li></ul>
  • 13. Content Delivery Networks P R1 R2 R3 C C C C C C C C C
  • 14. P2P Applications <ul><li>Many P2P applications since the 1990s </li></ul><ul><ul><li>File sharing </li></ul></ul><ul><ul><ul><li>Napster, Gnutella, KaZaa, BitTorrent </li></ul></ul></ul><ul><ul><li>Internet telephony </li></ul></ul><ul><ul><ul><li>Skype </li></ul></ul></ul><ul><ul><li>Internet television </li></ul></ul><ul><ul><ul><li>PPLive, CoolStreaming </li></ul></ul></ul>
  • 15. Why P2P? <ul><li>Every node is both a server and client </li></ul><ul><ul><li>Easier to deploy applications at endpoints </li></ul></ul><ul><ul><li>No need to build and maintain expensive infrastructure </li></ul></ul><ul><ul><li>Potential for both performance improvement and additional robustness </li></ul></ul><ul><ul><li>Additional clients create additional servers for scalability </li></ul></ul>
  • 16. P2P Multicast Purdue Stan1 Stan2 Berk2 Overlay Tree Stanford Berkeley Dumb Network Gatech Gatech Berk1 Stan1 Stan2 Berk1 Berk2 Source: Purdue Source: Sanjay Rao’s lecture from Purdue
  • 17. P2P Multicast <ul><li>Tree-based </li></ul><ul><ul><li>Parent-child relationships </li></ul></ul><ul><ul><li>Push-based </li></ul></ul><ul><ul><li>Uplink bandwidth not utilized at leaves </li></ul></ul><ul><ul><ul><li>Data can be divided and disseminated along multiple trees (e.g., SplitStream) </li></ul></ul></ul><ul><ul><li>Must be repaired and maintained to avoid interruptions </li></ul></ul><ul><ul><li>Example: End System Multicast (ESM) </li></ul></ul>
  • 18. P2P Multicast <ul><li>Mesh-based </li></ul><ul><ul><li>Data-driven </li></ul></ul><ul><ul><li>Pull-based </li></ul></ul><ul><ul><li>Periodically exchange data availability with random partners and retrieve new data </li></ul></ul><ul><ul><li>Unlike BitTorrent, must consider real-time constraints </li></ul></ul><ul><ul><li>Example: CoolStreaming </li></ul></ul>
  • 19. Overlay Performance <ul><li>Even a well-designed overlay cannot be as efficient as IP Mulitcast </li></ul><ul><li>But performance penalty can be kept low </li></ul><ul><li>Trade-off some performance for other benefits </li></ul>Increased Delay Source: Sanjay Rao’s lecture from Purdue Dumb Network Gatech Duplicate Packets: Bandwidth Wastage Stanford Berkeley
  • 20. Case Study: PPLive <ul><li>Very popular P2P IPTV application </li></ul><ul><ul><li>Free for viewers </li></ul></ul><ul><ul><li>Over 100,000 simultaneous viewers and 400,00 viewers daily </li></ul></ul><ul><ul><li>Over 200+ channels </li></ul></ul><ul><ul><li>Windows Media Video and Real Video format </li></ul></ul>
  • 21. Case Study: PPLive <ul><li>Gossip-based protocols </li></ul><ul><ul><li>Peer management </li></ul></ul><ul><ul><li>Channel discovery </li></ul></ul><ul><li>Data-driven p2p streaming </li></ul><ul><li>Recommend papers are measurement studies of PPLive </li></ul>
  • 22. Case Study: PPLive <ul><li>Contact channel server for available channels </li></ul><ul><li>Retrieve list of peers watching selected channel </li></ul><ul><li>Find active peers on channel to share video chunks </li></ul>Source: “Insights into PPLive: A Measurement Study of a Large-Scale P2P IPTV System” by Hei et al.

×