Redlarf (funded by Eurongi)

Redlarf

Research Project financed by

Redlarf Project Team

Heads

Ilkka Norros (VTT Technical Research Centre of Finland),

Kurt Tutschku (University of Wuerzburg),

Marco Mellia (Politecnico di Torino)

Researchers

Hannu Reittu,

Vesa Pehkonen,

Andreas Binzenhöfer,

Michela Meo,

Daniel Schlosser

Summary

BitTorrent introduced an extremely efficient peer-to-peer networking technique for distributing a very large file (say, gigabytes) to a very large number of recipients (scalability should extend to millions) by splitting it into relatively small chunks (say, thousand chunks) in such a way that the recipients forward their downloaded chunks to further recipients. This project studies fully distributed analogues on BitTorrent from several points of view: algorithm design, testing and analysis.

Main objectives:

Design and study, using analytical models, different mediation algorithms applicable in a file distribution system based on chunk transfer along random encounters. The objective is to maximize the file distribution performance in terms like throughput, delay and reliability.
Implement and test these algorithms in a working, experimental peer-to-peer content distribution system. Tests and performance measurements are made using partners' capabilities and, in particular, in the world-wide PlanetLab environment.
Analyse and test the reliability and stability performance of various overlay network alternatives of the system.

Meetings

20.11 - 22.11.2006 Meeting of partners from Würzburg and VTT in Helsinki
13.06 - 15.06.2006 Meeting of all three partners in Torino
07.05 - 13.05.2006 Visit of Vesa Pehkonen (VTT) to Würzburg
22.02 - 24.02.2006 Kick-off Meeting in Würzburg

Talks

22.11.2006, Helsinki: Andreas Binzenhöfer: REDLARF: Prototype Study in PlanetLab
13.06.2006, Torino: Daniel Schlosser: Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download
13.06.2006, Torino: Andreas Binzenhöfer: The REDLARF PlanetLab Management Envrionment

Papers

Ilkka Norros, Vesa Pehkonen, Hannu Reittu, Andreas Binzenhöfer, Kurt Tutschku
Relying on randomness - PlanetLab experiments based on random encounters
3rd EURO-NGI Conference on Next Generation Internet Networks (NGI 2007),
Trondheim, Norway, May 2007
Ilkka Norros, Balakrishna Prabhu, Hannu Reittu.
Flash crowd in a file sharing system based on random encounters
interperf '06, 2006
Daniel Schlosser, Tobias Hoßfeld, Kurt Tutschku.
Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download.
Technical Report No. 385, May 2006.

Related Work

E.W. Biersack, D. Carra, R. Lo Cigno, P. Rodriguez , P. Felber
Overlay Architectures for File Distribution: Fundamental Performance Analysis for
Homogeneous and Heterogeneous Cases
A. Legout, G. Urvoy-Keller, and P. Michiardi.
Rarest First and Choke Algorithms Are Enough
L. Nussbaum, O. Richard
Lightweight Emulation to Study Peer-to-Peer Systems

Publications

Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download Hoßfeld, Tobias (2006, September 7).
- [ BibTeX ]
- [ EndNote ]
@presentation{info3-talk-2006-224, address = {Sixth IEEE International Conference on Peer-to-Peer Computing (P2P2006), Cambridge, UK}, author = {Hoßfeld, Tobias}, keywords = {redlarf}, month = 9, title = {Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download}, year = 2006 }

%0 Generic %1 info3-talk-2006-224 %A Hoßfeld, Tobias %C Sixth IEEE International Conference on Peer-to-Peer Computing (P2P2006), Cambridge, UK %D 2006 %T Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download
Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download Schlosser, Daniel; Hoßfeld, Tobias; Tutschku, Kurt (2006).
- [ BibTeX ]
- [ EndNote ]
- [ Download ]
@techreport{info3-techreport-2006-196, author = {Schlosser, Daniel and Hoßfeld, Tobias and Tutschku, Kurt}, institution = {University of Wuerzburg}, keywords = {redlarf}, month = 5, number = 385, title = {Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download}, year = 2006 }

%0 Report %1 info3-techreport-2006-196 %A Schlosser, Daniel %A Hoßfeld, Tobias %A Tutschku, Kurt %D 2006 %N 385 %T Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download
Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download Schlosser, Daniel; Hoßfeld, Tobias; Tutschku, Kurt in Sixth IEEE International Conference on Peer-to-Peer Computing (P2P2006) (2006).
- [ Abstract ]
- [ BibTeX ]
- [ EndNote ]
- [ Download ]
The performance of Peer-to-Peer (P2P) content distribution networks depends highly on the coordination of the peers. Sophisticated cooperation strategies, such as the multiple source download, are the foundation for efficient file exchange. The detailed performance of the strategies are determined by the peer characteristics and the peer behaviour, such as the number of parallel upload connections, the selfishness, or the altruistic re-distribution of data. The purpose of this work is to evaluate and investigate different cooperation strategies for multiple source download and select the best one for a scenario for even leeching peers, i.e. peers which depart as soon as they have finished their download. The question arises whether the cooperation strategy can smoothen the overall performance degradation caused by a selfish peer behaviour. As performance indicator the evolution of the numbers of copies of a chunk and the experienced download times of files is applied. The considered scenarios comprise best-case (altruistic peers) and worst-case scenarios (selfish peers). We further propose a new cooperation strategy to improve the file transfer even when mainly selfish peers are present, the CygPriM (cyclic priority masking) strategy. The strategy allows an efficient P2P based content distribution using ordered chunk delivery with only local information available at a peer.

@inproceedings{info3-inproceedings-2006-239, abstract = {The performance of Peer-to-Peer (P2P) content distribution networks depends highly on the coordination of the peers. Sophisticated cooperation strategies, such as the multiple source download, are the foundation for efficient file exchange. The detailed performance of the strategies are determined by the peer characteristics and the peer behaviour, such as the number of parallel upload connections, the selfishness, or the altruistic re-distribution of data. The purpose of this work is to evaluate and investigate different cooperation strategies for multiple source download and select the best one for a scenario for even leeching peers, i.e. peers which depart as soon as they have finished their download. The question arises whether the cooperation strategy can smoothen the overall performance degradation caused by a selfish peer behaviour. As performance indicator the evolution of the numbers of copies of a chunk and the experienced download times of files is applied. The considered scenarios comprise best-case (altruistic peers) and worst-case scenarios (selfish peers). We further propose a new cooperation strategy to improve the file transfer even when mainly selfish peers are present, the CygPriM (cyclic priority masking) strategy. The strategy allows an efficient P2P based content distribution using ordered chunk delivery with only local information available at a peer.}, address = {Cambridge, UK}, author = {Schlosser, Daniel and Hoßfeld, Tobias and Tutschku, Kurt}, booktitle = {Sixth IEEE International Conference on Peer-to-Peer Computing (P2P2006)}, keywords = {redlarf}, month = 9, title = {Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download}, year = 2006 }

%0 Conference Paper %1 info3-inproceedings-2006-239 %A Schlosser, Daniel %A Hoßfeld, Tobias %A Tutschku, Kurt %B Sixth IEEE International Conference on Peer-to-Peer Computing (P2P2006) %C Cambridge, UK %D 2006 %T Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download %X The performance of Peer-to-Peer (P2P) content distribution networks depends highly on the coordination of the peers. Sophisticated cooperation strategies, such as the multiple source download, are the foundation for efficient file exchange. The detailed performance of the strategies are determined by the peer characteristics and the peer behaviour, such as the number of parallel upload connections, the selfishness, or the altruistic re-distribution of data. The purpose of this work is to evaluate and investigate different cooperation strategies for multiple source download and select the best one for a scenario for even leeching peers, i.e. peers which depart as soon as they have finished their download. The question arises whether the cooperation strategy can smoothen the overall performance degradation caused by a selfish peer behaviour. As performance indicator the evolution of the numbers of copies of a chunk and the experienced download times of files is applied. The considered scenarios comprise best-case (altruistic peers) and worst-case scenarios (selfish peers). We further propose a new cooperation strategy to improve the file transfer even when mainly selfish peers are present, the CygPriM (cyclic priority masking) strategy. The strategy allows an efficient P2P based content distribution using ordered chunk delivery with only local information available at a peer.

Hubland Süd, Geb. M2