Close

Performance Analysis of Updating Mechanisms for Dynamic Content in Peer-to-Peer Networks

Abstract

In current peer-to-peer applications content is generally static, i.e. it does not change over time. In the case of content that is modified over time, peers that host such dynamic content are required to have a mechanism that refreshes content hosted by them. In unstructured peer-to- peer networks this is challenging because of its decentralized nature. In this paper we develop a model that captures for dynamic content in a peer-to-peer network the impact of its refreshing rate and its popularity to the bandwidth required from the network. We compare four simple mechanisms for updating peers that can be used in a protocol devised to work in this kind of scenario. We find that there is an optimal number of updates to proactively push into the network in order to minimize the bandwidth consumption. Since unstructured peer-to-peer networks offer little guarantees of reliability, the use of one of the mechanisms studied here is highly recommended.

MIDAS Network Members

Citation: