An Evaluation Criterion for Adaptive Neighbor Selection in Heterogeneous Peer-to-Peer Networks

The peer-to-peer paradigm potentially enables low-cost and highly scalable distributed systems where user nodes have at the same time the roles of consumer and provider of resources. Nowadays, the computational power and storage capacity of mobile devices has notably increased, for which their inclusion in the physical network that supports P2P overlays can be realistically considered. For those applications where the quality of service is a fundamental constraint, such as live streaming, it is important to have overlay network level strategies to dynamically re-configure the active connections among peers. In this paper we propose a simple but effective strategy for the evaluation of neighbors, taking into account their capabilities, to support the dynamic selection of resource providers. We apply the formal framework to P2P Internet Live Streaming, simulating three realistic scenarios and discussing the results.

[1]  Srinivas Aluru,et al.  High Performance Computing - HIPC 2007 , 2008 .

[2]  Bo Li,et al.  Coolstreaming: Design, Theory, and Practice , 2007, IEEE Transactions on Multimedia.

[3]  Toufik Ahmed,et al.  P2P-based mobile IPTV: Challenges and opportunities , 2008, 2008 IEEE/ACS International Conference on Computer Systems and Applications.

[4]  Tomoya Enokido,et al.  A Multi-Source Streaming Model for Mobile Peer-to-Peer (P2P) Overlay Networks , 2008, ICDCS Workshops.

[5]  Ajith Abraham,et al.  Multi-objective Peer-to-Peer Neighbor-Selection Strategy Using Genetic Algorithm , 2007, HiPC.

[6]  Tomoya Enokido,et al.  A Multi-Source Streaming Model for Mobile Peer-to-Peer (P2P) Overlay Networks , 2008, 2008 The 28th International Conference on Distributed Computing Systems Workshops.

[7]  Olivier Fourmaux,et al.  Source vs Data-driven Approach for Live P2P Streaming , 2006, International Conference on Networking, International Conference on Systems and International Conference on Mobile Communications and Learning Technologies (ICNICONSMCL'06).

[8]  Simon G. M. Koo,et al.  On neighbor-selection strategy in hybrid peer-to-peer networks , 2006, Future Gener. Comput. Syst..

[9]  Danny H. K. Tsang,et al.  Building Heterogeneous Peer-to-Peer Networks: Protocol and Analysis , 2008, IEEE/ACM Transactions on Networking.

[10]  Michele Amoretti,et al.  DEUS: a discrete event universal simulator , 2009, SimuTools.