|
Dr
Carey Williamson
Wireless Traffic Modeling and
Simulation
Dr Carey Williamson has been awarded an
iCORE/TELUS
Mobility
Chair in Wireless Traffic Modelling and
Simulation. Dr Carey Williamson, also
an iCORE Professor in Broadband Wireless
Networks, will receive $100,000 a year
from both TELUS
Mobility
and iCORE for five years to undertake
research into the modeling and analysis
of wireless Internet traffic.
Biographical
Information
Dr Carey Williamson is considered one
of the "rising stars" of computer
networks research. He is a well-respected
member of the ACM SIGCOMM and ACM SIGMETRICS
research communities, and has made significant
research contributions to computer networks
and performance evaluation and workload
characterization of web servers. He is
particularly noted for demonstrating the
usefulness of loss-load curves as a network
congestion control mechanism, and a unique
protocol supporting IP multicast for mobile
hosts. His strong publication record is
reflected in over 50 refereed publications
(conferences and journals), sixteen completed
graduate student theses and numerous technical
reports and software artifacts. He is
Chair of the IEEE Computer Society Chapter
of the IEEE North Saskatchewan Section,
and Secretary Treasurer of ACM SIGMETRICS.
Dr. Williamson has also been voted "Professor
of the Year" by graduating Computer
Science undergraduate students four times.
He also won the Master Teacher award at
the University of Saskatchewan Fall Convocation
in October 2000. Dr Williamson joins the
University of Calgary from the Department
of Computer Science at the University
of Saskatchewan.
Research Program
Research goals include developing and
validating traffic models for different
wireless service classes such as voice,
email, images, video, multimedia and web
access. The goal of the research is to
support the design and provisioning of
wireless and wired backbone networks.
On
voice services the traditional method
for provisioning is to model the number
of channels versus occupancy of the channels,
in order to maintain a blocking grade
of service. Voice traffic intensity is
characterized by specifying the number
of busy hour call attempts and the average
holding time per call.
Then,
using a model for a given blocking probability,
the number of channels required per sector
is determined. In some types of data transfer,
this method works, assuming an average
power per user going out, and an average
pole capacity and loading on the reverse
path; all three factors need to be determined
per sector.
TELUS
Mobility assumes voice and data traffic
can be modelled separately in a mutually
exclusive environment. Therefore there
exists a need for a method to model fundamental
and supplemental data channels in opposition
to the occupancy of the channels to maintain
data throughput, or packet delay probability.
An
initial three-step provisioning methodology
will give TELUS Mobility an efficient
plan for network growth leading to competitive
service level. These initial steps are
to:
- Develop
a method for characterizing data traffic
intensity for specific data services
and the combination of these services;
- Develop
a rule of thumb occupancy versus channels,
for maintaining a data throughput;
- Define
the range of accuracy of the rule of
thumb and factors having the largest
impact.
Related
Links:
Dr
Williamson's Homepage
|
