1 / 18

BiToS: Enhancing BitTorrent for Supporting Streaming Applications

BiToS: Enhancing BitTorrent for Supporting Streaming Applications. Aggelos Vlavianos Marios Iliofotou Michalis Faloutsos University of California Riverside. What is BiToS. BiToS = BitTorrent Streaming A BT-based protocol with the ability to support streaming

manny
Download Presentation

BiToS: Enhancing BitTorrent for Supporting Streaming Applications

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. BiToS: Enhancing BitTorrent for SupportingStreaming Applications Aggelos VlavianosMarios Iliofotou Michalis Faloutsos University of California Riverside

  2. What is BiToS • BiToS = BitTorrent Streaming • A BT-based protocol with the ability to support streaming • Modifies the piece selection policy of BT and nothing else • Uses a variable size sliding time window

  3. BitTorrent • I skipped this part

  4. Related Work • Bass (2005) • Hybrid server/P2P streaming system • Uses an almost unmodified version of BT • CoolStreaming • Uses a fixed sliding time window • Chainsaw • Uses gossip and push-based approaches

  5. BitTorrent Limitations • Piece selection mechanism • Based on rarity • Neglects deadlines • Poorly suited to Live Video Streaming • Would have to redesign the tracker • Can be tuned to support Playback Video Streaming

  6. BiToS

  7. Received pieces • State of a piece can be • Downloaded • Not-Downloaded • Missed (too late!)

  8. High priority set • Contains pieces that are • Not downloaded yet • Not missed • Close to be reproduced by the player • Set has a fixed size • BiToS parameter

  9. Remaining pieces set • Contains pieces that are • Not Downloaded • Not Missed • Not in the High Priority Set

  10. Selection process • Peer select with probability • p to select a piece in its High Priority Set • 1 – p to select a piece in the remaining pieces set • Piece selection is rarest first • When a piece from the High Priority Set is downloaded, it is replaced in that set by the first piece in sequence from the Remaining Pieces Set

  11. Timeliness of Arrival • Pieces that cannot meet playback deadlines are marked Missed

  12. The effect of probability p • p = 1 means that the peer will only download pieces that are in the High Priority Set • Peer will have the pieces it needs • Not clear that these pieces will be in high demand by other peers • Everyone is likely to have them • Peer may have problems later trading pieces with other peers

  13. The effect of probability p • p < 1 means that the peer will also download pieces that are not in its High Priority Set • Will be in the Remaining Pieces Set • Will be less likely to have been downloaded by too many peers • Can be traded later for other pieces

  14. Experimental evaluation • BitTorrent simulator • 4 seeders • 400 peers arriving in flash crowd • Ten minute video at 500 Kb/s • Upload and download rates are equal • 500 Kb/s

  15. Experimental results • Simulated three piece selection mechanisms • Sequential order within HPS and p = 1 • Rarest-first order within HPS and p = 1 • Rarest-first order within HPS and p = 0.8 • Use Continuity Index as benchmark

  16. Experimental results

  17. Experimental results

  18. Conclusions • Sliding time windows work! • Best sliding time window size is 5-10 percent of file size • Rarest-first policy outperforms sequential policy by a wide margin

More Related