RMIT University
Browse

Supporting disconnection operations through cooperative hoarding

Download (273.7 kB)
conference contribution
posted on 2024-11-23, 00:30 authored by Kwong Lai, Zahir TariZahir Tari, Peter Bertok
Mobile clients often need to operate while disconnected from the network due to limited battery life and network coverage. Hoarding supports this by fetching frequently accessed data into clients' local caches prior to disconnection. Existing work on hoarding have focused on improving data accessibility for individual mobile clients. However, due to storage limitations, mobile clients may not be able to hoard every data object they need. This leads to cache misses and disruption to clients' operations. In this paper, a new concept called cooperative hoarding is introduced to reduce the risks of cache misses for mobile clients. Cooperative hoarding takes advantage of group mobility behaviour, combined with peer cooperation in ad-hoc mode, to improve hoard performance. Two cooperative hoarding approaches are proposed that take into account access frequency, connection probability, and cache size of mobile clients so that hoarding can be performed cooperatively. Simulation results show that the proposed methods significantly improve cache hit ratio and provides better support for disconnected operations compared to existing schemes.

History

Related Materials

  1. 1.
    ISBN - Is published in 0780394283 (urn:isbn:0780394283)

Start page

35

End page

42

Total pages

8

Outlet

Conference on Computer Communications and Networks (ICCCN)

Editors

Y. Yang and S. Thuel

Name of conference

International Conference on Computer Communications and Networks

Publisher

IEEE

Place published

Piscataway, USA

Start date

2005-10-17

End date

2005-10-19

Language

English

Copyright

© 2005 IEEE

Former Identifier

2005001105

Esploro creation date

2020-06-22

Fedora creation date

2009-04-08

Open access

  • Yes

Usage metrics

    Scholarly Works

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC