|
Program Leader: Professor
Iven Mareels (University of Melbourne)
The Distributed Sensor and Information Systems Program’s research
ranges from the investigation of fundamental problems in algorithmic
design to the development and field testing of practical systems.
The research is applied to the measurement of performance in active
communication systems, the conservation of water in irrigation
networks and the performance of large-scale radar surveillance.
The Program comprises four research projects:
The Multi-Sensor Multi-Target Tracking Project and The Information
Fusion Project research the integration of data from a large number
of diverse sensors in order to develop situation assessments over
large geographic areas as required in wide area surveillance.
Brute force computational methods to solve such problems suffer
from the curse of dimensionality and their complexity scales exponentially.
Random algorithms have provided a way to break this linkage for
a wide class of problems and CSSIP has provided these new algorithms
and proof of their convergence in the context of large-scale surveillance.
CSSIP is now developing a generic software test-bed environment
to enable researchers to prototype algorithms and test them against
realistic scenarios.
This research receives strong support from DSTO through their
contribution to The Data Fusion Laboratory.
The Distributed Systems Design and Simulation Project
researches the conservation of water in large irrigation networks.
This is an important ecological and economic problem and shows
how a distributed systems design approach can deliver great economic
and social benefit. A conservative estimate indicates that about
20 to 30% of the dispatched water in irrigation canals world wide
does not reach farm land and is lost for irrigation purposes.
By exploiting mathematical models of irrigation canals and implementing
computer controlled closed loop control strategies these losses
can be substantially reduced.
This project is supported by AusIndustry and RLM Systems Pty
Limited with hardware for the remote sensing of water levels,
water flows and remote automated gate control (actuators in the
canals) provided by Rubicon Systems Pty Limited
CSSIP is developing the software for the models, control strategies
and scheduling. Algorithms have been tested in the field with
performance targets being met. Development work to implement and
engineer the proposed solutions into a robust, user friendly software
platform is in progress and an active technology transfer program
is ensuring that end-users are informed and enabled to develop
state-of-the-art designs.
Current issues being addressed are in the difficult areas of
demand forecasting and supply prediction. Both have an important
impact on the optimal scheduling of water resources.
The Performance Analysis Project researches the performance
of active networks which are communication networks governed by
programmable routers that are able to execute traffic dependent
code and promise the development of more flexible and robust networks.
Such flexibility comes with the inherent difficulty of predicting
network behaviour and CSSIP’s research concentrates on understanding
network behaviour and predicting quality of service performance
measures.
The project is developing an understanding of scaling issues,
their relationship to the design of active networks and the prediction
of performance for complex interconnections of subsystems. The
simulation of large-scale systems with the aim of providing meaningful
performance estimates is a difficult problem with random algorithms
currently providing the best approach to break down the computational
complexity issues.
For further details please contact:
Mr Geoff Vaughan-Evans
Centre Manager
CSSIP
Building P, Mawson Lakes Campus
University of South Australia
MAWSON LAKES
SA 5095 Australia
Phone: +61 8 8302 3923
Fax: +61 8 8302 5301
Email: gve@cssip.edu.au
or
Professor Iven Mareels
Department of Electrical and Electronic Engineering
The University of Melbourne
Parkville
VIC 3010
Phone: +61 3 8344 6699
Fax: +61 3 8344 7412
Email: i.mareels@ee.mu.oz.au
http://www.ee.mu.oz.au
|
