Effect of Prediction Errors on Adaptive Modulation Systems for Wireless Channels.

Sorour Falahati, Chalmers U. of Technology,
Arne Svensson, Chalmers U. of Technology,
Torbjörn Ekman, Uppsala University and
Mikael Sternad, Uppsala University

Radiovetenskap och Kommunikation RVK 02, Stockholm, June 2002.

Outline:

The Wireless IP project within the PCC program studies problems that are crucial in the evolution of UMTS towards high data rates, as well as in future 4G technologies aimed at rapidly mobile terminals. The goal is to attain higher througputs for packet data in particular in downlinks, without bandwidth expansion and while providing acceptable quality of service for various classes of traffic.

Abstract:

We study the optimum design of an adaptive modulation scheme based on M-QAM modulation assisted by channel prediction for the flat Rayleigh fading channel. The data rate and transmit power are adaptedto maximize the spectral efficiency, subject to average power and instantaneous BER constraints.

Optimum solutions for the rate and transmit power are derived based on the predicted SNR as well as the prediction error variance which are assumed to be available at the transmitter. The analytical result are evaluated and presented.

Related publications:

Journal paper version, (IEEE TCOM 2004) with further results.

PhD Thesis on channel prediction by Torbjörn Ekman, Oct. 2002.

VTC02-fall paper on channel prediction over 0.1-0.8 wavelengths.

A system proposal in which predictive adaptive modulation is used (RVK02).

An overview of the Wireless IP Project, RVK02.

Source:

Pdf, (228K)
Postscript, (208K)
Poster, in Pdf (660K)