The Wireless IP Project
Radiovetenskap och Kommunikation RVK 02,
Stockholm, June 2002.
The Swedish Foundation of Strategic Research funds the research
program PCC since
1997. More than 40 PhD students, 30 advisors and senior researchers and several
industrial partners are and organized in six subprojects at
four major Swedish universities.
The aim is to work towards personal multimedia communication
to all, at the same cost as
fixed telephony today.
In particular we study technologies for fourth generation (4G)
wireless systems. Here the
wireless IP project develops innovative approaches to increase
spectrum efficency and
througput for packet data over wireless links.
The optimization of resources in
wireless packet data systems
is challenging when users are mobile
with time-varying link quality. Within the
SSF PCC program, we have since year 2000
formed the Wireless IP project, which
studies such issues.
We perform research towards a flexible
packet data system with
wide area coverage for mobile users.
This system should have
performance equivalent to a 100Mbit/s Ethernet
connection, with support for Quality of Service and
fairness between wireless users.
This will require an optimization
that involves many levels in the OSI stack.
In the project, we develop and
integrate a set of key technologies, consisting
of prediction of radio channels, fast link adaptation,
scheduling of several packet data streams
over time-varying channels,
fast resource allocation using antenna arrays
and the use of information from lower layers in
higher layers, in particular the transport layer.
With these methods, the transmission
of packet data to/from mobile users
can be radically improved,
without expensive use of additional spectral bandwidth.
The downlink of our OFDM target
system, with high spectral efficiency.
Licenciate Thesis by Nilo C. Ericsson
on predictive scheduling combined with adaptive modulation.
Licenciate Thesis by Sorour Falahati, Feb. 2000.
on channel prediction over 1-10 ms horizons.
Scheduling and adaptive transmission
for the downlink 4G systems similar to the one discussed here
(FTC 2001, Beijing).
An overview of the Wireless IP Project, NRS 01.
Some other presentations by Wireless IP at RVK02:
Improved Channel Power Prediction,
by using unbiased predictors and advanced regressor noise reduction.
Optimizing adaptive modulation,
by taking the power prediction error into account.
Simulating TCP/IP traffic
over wireless links in real time.