Artificial neural network approach for MMIC passive and active device characterization

Giannini, F. ; Leuzzi, G. ; Orengo, G. ; Albertini, M. (2000) Artificial neural network approach for MMIC passive and active device characterization. In: Gallium Arsenide applications symposium. GAAS 2000, 2-6 october 2000, Paris.
Full text available as:
[thumbnail of GAAS12_5.pdf]
Preview
PDF
Download (166kB) | Preview

Abstract

Artificial neural networks (ANNs)are presented for the fast and accurate modeling of passive and active devices in MMICs.ANNs trained with S-parameter over a wide range of frequencies were used for rectangular spiral inductor modeling and HEMT device small-signal models.After training and testing the ANN S-parameter model can be implemented as a two-port network into commercial circuit simulators.This enable the ANN model to be used in the design,analysis,and optimization of microwave circuits.A combination of ANN models trained with S-parameter and dc drain current measurements was implemented into an HP microwave circuit simulator to provide a small- signal bias-dependent active model.The proposed technique is capable of providing easy,fast and accurate simulation models for MMIC components where models based on equivalent circuit parameters (ECPs)do not exist or are not accurate over the desired region of operation.

Abstract
Document type
Conference or Workshop Item (Paper)
Creators
CreatorsAffiliationORCID
Giannini, F.
Leuzzi, G.
Orengo, G.
Albertini, M.
Subjects
DOI
Deposit date
17 Jun 2004
Last modified
17 Feb 2016 13:41
URI

Other metadata

Downloads

Downloads

Staff only: View the document

^