Trace Analysis of a Wireless University Network with Authentication

Alexander  Pelov; Pierre David; Thomas Noël

1st International ICST Workshop On Wireless Network Measurement

Research Article

Trace Analysis of a Wireless University Network with Authentication

Cite: BibTeX Plain Text

@INPROCEEDINGS{10.1109/WIOPT.2007.4480111,
    author={Alexander  Pelov and Pierre David and Thomas No\`{\i}l},
    title={Trace Analysis of a Wireless University Network with Authentication},
    proceedings={1st International ICST Workshop On Wireless Network Measurement},
    publisher={IEEE},
    proceedings_a={WINMEE/WITMEMO},
    year={2008},
    month={3},
    keywords={Authentication  Data analysis  IP networks  Network servers  Performance analysis  Personal digital assistants  Portals  Telecommunication traffic  Web server  Wireless networks},
    doi={10.1109/WIOPT.2007.4480111}
}

Alexander Pelov
Pierre David
Thomas Noël
Year: 2008
Trace Analysis of a Wireless University Network with Authentication
WINMEE/WITMEMO
IEEE
DOI: 10.1109/WIOPT.2007.4480111

Alexander Pelov¹^,*, Pierre David²^,*, Thomas Noël³^,*

1: LSIIT (UMR7005) – University Louis Pasteur, blvd Sébastien Brant, 67400 Illkirch, France (phone: +33390244587; fax: +33390244455)
2: Center of Network Communications (CRC) of ULP - University Louis Pasteur, 7 rue René Descartes, 67084 Strasbourg Cedex, France
3: LSIIT (UMR7005) – University Louis Pasteur, blvd Sébastien Brant, 67400 Illkirch, France

*Contact email: pelov@dpt-info.u-strasbg.fr, Pierre.David@crc.u-strasbg.fr, Thomas.Noel@dpt-info.u-strasbg.fr

Abstract

In this paper we have analyzed traces gathered from 36 buildings part of 3 different universities, libraries, laboratories and public places. The availability of login information allowed us to quantify the MAC address sharing. In addition, we have described a procedure applicable to both WPA/802.1X and captive portal authentication logs, which in conjunction with association traces can be used to uniquely identify users with increased confidence. We found out that a non-negligible percentage - 36% - of the 400 thousand associations we studied are without corresponding authentications. We also found that 32% of the users have used the secured network and have caused 45% of the associations.

Keywords: Authentication Data analysis IP networks Network servers Performance analysis Personal digital assistants Portals Telecommunication traffic Web server Wireless networks

Published: 2008-03-31
Publisher: IEEE
Modified: 2011-07-31

: http://dx.doi.org/10.1109/WIOPT.2007.4480111