Fan Bai (mobility modeling
and IMPORTANT) [graduated with
PhD. Now at GM labs]
[NOTE: This web page is out-of-date! Please point to the
following link
at the University of Florida for updated information
http://nile.cise.ufl.edu/MobiLib].
This webpage aims to establish a community-wide library of mobile wireless
networks traces and measurements.
The goal is to have the traces, simulation code,
test-suites and models (of mobility, traffic and user behavior)
established by the experts in the field, widely available
for everyone to use and compare against.
Among the major universities that agreed to provide traces
are USC, MIT, UCSD, Dartmouth, UCSB, UIUC, Georgia Tech,
Purdue, UCLA, Rice, Boston U, Columbia,
U Washington, UNC. Links are continuously
updated.
If interested and can contribute
contact helmy@usc.edu
* New * W. Hsu, A. Helmy, "IMPACT:
Investigation of
Mobile-user Patterns Across University
Campuses using WLAN Trace Analysis", USC Technical
Report
(under submission), July 2005.
[Longer
technical report (update July 18, 05).
IMPACT webpage]. [This paper includes most information in the 2
Winmee papers above. To be extended further in the future.]
Highlights:
This paper provides the most comprehensive study of WLAN traces to date.
Traces collected from four major universities (~12,000 users) are analyzed
using metrics for individual user and group behaviors. Similarities and
differences across campuses are studied. Conclusions provide great insight
into realistic behavior of wireless users. Most users are 'on' for a small
fraction of the time, number of access points visited (per user) is quite
low, and on-line user mobility is quite low. On average, a user encounters
only 2%-6% of the user population. Encounter-graphs and small worlds are
introduced to model encounter patterns between users. We find that number
of encounters follows a biPareto distribution and the frienship indexes
follow exponential distributions. A paradigm for 'encounter-based
information diffusion' is introduced for efficient data dissemination in
mobile networks.
* New: Mar 28, 06*
W. Hsu, A. Helmy, "Principal Component Analysis of User Association
Patterns in Wireless LAN Traces", IEEE INFOCOM poster, April
2006.
* New: Mar 28, 06*
W. Hsu, A. Helmy, "Capturing User Friendship in WLAN Traces", IEEE
INFOCOM poster, April 2006.
* New * F. Bai, A. Helmy,
"Impact
of Mobility on Mobility-Assisted
Information Diffusion (MAID) Protocols", USC Technical Rreport,
July 2005. [Submitted for review] [Longer, more
complete, 29-page version] Highlights: This study analyzes a class of protocols, MAID, that
utilize mobility for information diffusion.
MAID uses encounter
information to create age gradients towards the target, and can be used
for discovering resources, routing or
locating nodes efficiently in future mobile networks.
Analytical models are developed to evaluate MAID's performance during its
various (transient and steady-state) phases of operation. Extensive
simulations are used to validate these models and to study the
sensitivity of MAID to a rich set of mobility models.
We find that although MAID is sensitive to the mobility pattern, its
steady state performance is, surprisingly, insensitive to velocity.
We identify the properties of the 'age gradient tree' as the key factor to
explain this interplay between mobility and the MAID protocols.
Thanks to the people who have contributed traces and/or encouraged this
work:
Mostafa Ammar, Richard Fujimoto (Georgia Tech),
Kevin Almeroth, Elizabeth Royer (UCSB),
David Kotz, Andrew Campbell (Dartmouth),
Nitin Vaidya, Jennifer Hou (UIUC),
Ness Schroff, Sonia Fahmy (Purdue), Mario Gerla, Medy
Sanadidi (UCLA),
Tracy Camp (Colorado School of
Mines), David Wetherall (U. Washington), Victor Bahl (Microsoft
Research), Ed Knightly, David Johnson (Rice),
Rene Cruz (UCSD), Maria Papadopouli, Kevin Jeffay (U North Carolina),
Henning Schulzrinne
(Columbia), Azer Bestavros, Ibrahim Matta (Boston U), Dina Katabi (MIT),
Stefano Basagni (Northeastern U.), Michele Zorzi (U. Padova/UCSD), Eylem
Ekici (Ohio State U),
Jim Kurose, Brian Levine (U. Mass - Amherst), Srikanth Krishnamurthy,
Michalis Faloutsos (UC Riverside)
This material is based upon work supported in part by the National Science
Foundation under Grant No. 0134650.
Any opinions, findings and conclusions or recomendations expressed in this
material are those of the author(s) and do not necessarily reflect the
views of the National Science Foundation (NSF).
Number of visitors since Jul. 18 2005:
This page was created May '05. Updated continuously...
