Home // ICNS 2011, The Seventh International Conference on Networking and Services // View article

Deploying a High-Performance Context-Aware Peer Classification Engine

Authors:
Mircea Bardac
George Milescu
Adina Magda Florea

Keywords: classification; BitTorrent context-awareness; peer-to-peer

Abstract:
This research presents a generic approach for context-aware entity classification with emphasis on integration and use of contextual information in Peer-to-Peer systems. The designed peer classification engine isolates high-latency update processes in order to minimize the latencies of the lookup queries. By using a key-value data-store with support for sorted sets, high complexity context-classifying functions can be executed asynchronously without impacting the lookup queries. The performance of the system is evaluated through experimental and complexity analysis, identifying directions for improving and scaling the peer classification engine. As ubiquitous computing evolves and becomes part of everyday life, the designed context- aware classification engine provides a basis for deploying the next-generation network-based services.

Pages: 268 to 273

Copyright: Copyright (c) IARIA, 2011

Publication date: May 22, 2011

Published in: conference

ISSN: 2308-4006

ISBN: 978-1-61208-133-5

Location: Venice/Mestre, Italy

Dates: from May 22, 2011 to May 27, 2011