On the Role of Helper Peers in P2P Networks

Recent studies in peer-to-peer (P2P) networks present surprising new designs that rely on helper peers. Helper peers, sometimes named as Feeders or Support peers are nodes that do not function as direct consumers or providers of content but are used to collaborate with other peers in the network for a growing variety of benefits. In File Sharing networks for instance, due to frequent joins, leaves and the characteristic fluctuating throughput of source peers, clients usually download at an unstable rate. In addition, existing P2P protocols tend to ignore source peers that have relatively low bandwidth to offer and practically miss a potentially huge resource. By employing helper peers that are optimal for availability and throughput stability with the downloading client, it is possible to provide a maximal stable throughput even with extremely weak and unstable sources. Other interesting examples of helper peers in file sharing demonstrated how to integrate helper peers in order to increase the number of sources under flash crowds situations, how to solve the last chunk problem and how to bypass fairness rules for better download rates. In P2P streaming networks such as live IPTV and VOD, helper peers can contribute in preventing glitches and expanding the dissemination of packets, as well as synchronizing and ordering frames for the clients. In this chapter we present novel architectures that embed helper peers in order to solve key problems in P2P networks. We discuss the implications and key techniques in each proposal and point the weaknesses and limitations of mentioned architectures. We present different selection criteria for choosing the optimal helper peers based on theoretic simulations, practical measurements and experiments with popular protocols such as eMule and BitTorrent. We propose an advanced Machine Learning based design that actively learns the behavioural patterns of peers and leverages the performance of clients by collaborating with the ”right” helper peers at the right time. Though helper peers gained popularity in P2P research, different works in this field term the same ideas differently and in some cases do not mention each other; this chapter presents the current state of the art in helper-supported P2P networks. Finally, we present future research directions in this field. 11

[1]  Jun Wang,et al.  TRIBLER: a social‐based peer‐to‐peer system , 2008, IPTPS.

[2]  Alexandru Iosup,et al.  2Fast : Collaborative Downloads in P2P Networks , 2006, Sixth IEEE International Conference on Peer-to-Peer Computing (P2P'06).

[3]  Danny Dolev,et al.  Collabory: A Collaborative Throughput Stabilizer & Accelerator for P2P Protocols , 2008, 2008 IEEE 17th Workshop on Enabling Technologies: Infrastructure for Collaborative Enterprises.

[4]  Tobias Hoßfeld,et al.  Comparison of Robust Cooperation Strategies for P2P Content Distribution Networks with Multiple Source Download , 2006, Sixth IEEE International Conference on Peer-to-Peer Computing (P2P'06).

[5]  Antonio Ortega,et al.  PALS: peer-to-peer adaptive layered streaming , 2003, NOSSDAV '03.

[6]  Kien A. Hua,et al.  P2VoD: providing fault tolerant video-on-demand streaming in peer-to-peer environment , 2004, 2004 IEEE International Conference on Communications (IEEE Cat. No.04CH37577).

[7]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[8]  Indranil Gupta,et al.  Mapping the PPLive Network: Studying the Impacts of Media Streaming on P2P Overlays , 2006 .

[9]  Rafit Izhak-Ratzin,et al.  Collaboration in BitTorrent Systems , 2009, Networking.

[10]  Keith W. Ross,et al.  A Measurement Study of a Large-Scale P2P IPTV System , 2007, IEEE Transactions on Multimedia.

[11]  Nazareno Andrade,et al.  On the Efficiency and Cost of Introducing QoS in BitTorrent , 2007, Seventh IEEE International Symposium on Cluster Computing and the Grid (CCGrid '07).

[12]  B. Cohen,et al.  Incentives Build Robustness in Bit-Torrent , 2003 .

[13]  Pei Cao,et al.  Can self-organizing P2P file distribution provide QoS guarantees? , 2006, OPSR.

[14]  Danny Dolev,et al.  Collabrium: Active Traffic Pattern Prediction for Boosting P2P Collaboration , 2009, 2009 18th IEEE International Workshops on Enabling Technologies: Infrastructures for Collaborative Enterprises.

[15]  Joseph Hao,et al.  Enhancing Collaborative Content Delivery with Helpers , 2004 .

[16]  Vinod M. Prabhakaran,et al.  On the Role of Helpers in Peer-to-Peer File Download Systems: Design, Analysis and Simulation , 2007, IPTPS.

[17]  Francisco de Asís López-Fuentes,et al.  Multi-source video multicast in peer-to-peer networks , 2008, 2008 IEEE International Symposium on Parallel and Distributed Processing.

[18]  Danny Dolev,et al.  Maxtream: Stabilizing P2P Streaming by Active Prediction of Behavior Patterns , 2009, 2009 Third International Conference on Multimedia and Ubiquitous Engineering.

[19]  Xiaoning Ding,et al.  Measurements, analysis, and modeling of BitTorrent-like systems , 2005, IMC '05.

[20]  Ashwin Machanavajjhala,et al.  P-Ring: An Index Structure for Peer-to-Peer Systems , 2004 .

[21]  Kannan Ramchandran,et al.  Enhancing peer-to-peer live multicast quality using helpers , 2008, 2008 15th IEEE International Conference on Image Processing.

[22]  Thomas E. Anderson,et al.  One Hop Reputations for Peer to Peer File Sharing Workloads , 2008, NSDI.

[23]  Fernando Cores,et al.  DynaPeer: A Dynamic Peer-to-Peer Based Delivery Scheme for VoD Systems , 2007, Euro-Par.