Behavioural Modelling of Radio Frequency Power Amplifiers.
An Evaluation of Some Block Structure and Neural Network Models.
Licentiate Thesis, Signals and Systems,
Uppsala University, December 2005.
Licentiate Seminar Friday December 16, 2005, 10.15 - 12.00
Room Översten, Building Grenadjären, Gävle Technology Park,
The thesis will be available from Signals and Systems, Uppsala University,
Box 534, SE-751 21 Uppsala, Sweden.
This work considers behavioral modelling of radio frequency power
amplifiers. Due to the
use of modern digital modulation methods power amplifiers are nowadays
signals having a considerable bandwidth and a fast changing envelope.
This means that
traditional quasi-memoryless amplitude-to-amplitude (AM/AM) and
(AM/PM) characteristics are no longer enough to describe and model the
power amplifiers, neither can they be successfully used for linearization.
In this thesis, sampled input and output data are used for
identification and validation of
some block structure models with memory. The time-discrete Volterra
model, the Wiener
model, the Hammerstein model, and the radial-basis function neural
network are all
identified and compared with respect to in-band and out-of-band errors.
Two different signal
types, i.e. multi tones and noise, with different powers, peak-to-average
bandwidths have been used as input to the amplifier. Two different power
investigated, one designed for the third generation mobile
telecommunication systems and
one for the second generation.
A stepped three-tone measurement technique based on digitally modulated
is presented. The third-order Volterra kernel were determined from
inter-modulation products. The properties of the Volterra kernel along
certain parts in the
three dimensional frequency space were analysed and compared to the