Giovanna Carofiglio
Abstract
We apply basic concepts of statistical physics to devise an approximate model describing the dynamics of content diffusion in large peer-to-peer networks. Our approach is based on fluid-diffusive equations, whose solution can be obtained by numerical evaluation with a complexity independent of the number of users and contents, thus allowing to analyze very large systems. The model is general and modular, and can incorporate the effect of both search and download processes.
- D. Qiu, R. Srikant, "Modeling and Performance Analysis of BitTorrent-Like Peer-to-Peer Networks", ACM SIGCOMM 2004, USA, 2004 Google Scholar
Digital Library
- W.Whitt, A Diffusion Approximation for the G/GI/n/m Queue. Operations Research, vol. 52, No. 6, Nov.-Dec. 2004, pp. 922--941. Google ScholarDigital Library
- R. Gaeta, G. Balbo, S Bruell, M. Gribaudo, M. Sereno. A simple analytical framework to analyze search strategies in large-scale peer-to-peer networks. In Proc. of Performance 2005 Conference, October 2005, Juan-les-Pins, France.Google ScholarDigital Library
Index Terms
- A statistical physics approach for modelling P2P systems
Recommendations
Modeling Churn in P2P Networks
ANSS '07: Proceedings of the 40th Annual Simulation SymposiumThe objective of this paper is to introduce a model to guide the analysis of the impact of churn in P2P networks. Using this model, a variety of node membership scenarios is created. These scenarios are used to capture and analyze the performance trends ...
An Approach to Build a P2P Content Distribution Structure of High Performance
ICICSE '12: Proceedings of the 2012 Sixth International Conference on Internet Computing for Science and EngineeringIt is well known that content distribution networks (CDN) are expensive technologies for distributing web contents. Universities and other non-profit organizations do not have enough financial means to deploy large web-servers or to use content ...
Stochastic Optimization for Content Sharing in P2P Systems
Available resources in peer-to-peer (P2P) systems depend strongly on resource contributions made by individual peers. Empirical data shows that in the absence of incentives, a majority of the participating peers do not contribute resources. Modeling ...
Comments