Picture of scraped petri dish

Scrape below the surface of Strathprints...

Explore world class Open Access research by researchers at the University of Strathclyde, a leading technological university.

Explore

An optimum network selection solution for multihomed hosts using Hopfield Networks

Espi, Jorge and Atkinson, R.C. and Harle, D.A. and Andonovic, Ivan (2010) An optimum network selection solution for multihomed hosts using Hopfield Networks. In: Ninth International Conference on Networks, 2010. IEEE, pp. 249-254. ISBN 978-1-4244-6083-0

[img] PDF (hnn.pdf)
hnn.pdf
Restricted to Registered users only

Download (272kB) | Request a copy from the Strathclyde author

Abstract

This work introduces a Hopfield Neural Network approach to network selection for multihomed hosts which considers a range of relevant network parameters including available radio access technologies and traffic types (VoIP, video streaming, Web browsing and FTP-based). Also proposed is a novel utility function that further improves network selection. Results show that, in terms of QoS, the allocation obtained using proposed algorithm outperforms other two reference allocation schemes under a range of different scenarios.