Title :
High Power Amplifier Dynamic Modeling by Modified Volterra Series and Neural Network Tool
Author :
Dominique, C. ; Mancuso, Y. ; Dueme, P. ; Gautier, J-L.
Author_Institution :
THALES AIRBORNE SYSTEMS - La Clef de St Pierre - 1 bd Jean Moulin - 78852 Elancourt Cedex - France. Tel:+3301.34.59.68.00 - Fax:+3301.34.59.55.99 - cyril.dominique@fr.thalesgroup.com
Abstract :
Large signal system level simulation needs to build behavioural models. The black box approach is tailored for modulated excitation signals, especially in a non-linear system with memory, like Solid State Power Amplifier (SSPA), where the response depends both on power and frequency parameters. In this study, memory effects are taken into account by non-linear analytic functions based on neural network fit. Impacts on CW and intermodulation characteristics of non-linear memory are presented.
Keywords :
Artificial neural networks; Bandwidth; Costs; Frequency; High power amplifiers; Kernel; Microwave devices; Neural networks; Power system modeling; Scattering parameters;
Conference_Titel :
Microwave Conference, 2001. 31st European
Conference_Location :
London, England
DOI :
10.1109/EUMA.2001.339068
