Implementation of a Neural Network HEMT Model Into ADS

Neural network algorithms have been applied to various areas of engineering and computer science. A high electron-mobility transistor (HEMT) large-signal model has been implemented in the Advanced Design System (ADS) software package using a multilayered neural network. The neural network model was...

Full description

Saved in:
Bibliographic Details
Published inMicrowave Journal Vol. 44; no. 11; pp. 66 - 76+78
Main Authors Thompson, Willie L. II, Miller, Eric V, White, Carl
Format Journal Article Trade Publication Article
LanguageEnglish
Published Dedham Horizon House Publications, Inc 01.11.2001
EditionInternational ed.
Subjects
Accuracy
Algorithms
Back propagation
Colleges & universities
Computer engineering
Computer software industry
Design
Design and construction
High performance systems
Instrument industry
Mathematical functions
Microwave communications
Neural networks
Optimization
Physics
Product information
Propagation
Simulation
Software
Systems design
Transistors
United States
United States > US
Online AccessGet full text
ISSN0192-6225

Cover

More Information
Summary:Neural network algorithms have been applied to various areas of engineering and computer science. A high electron-mobility transistor (HEMT) large-signal model has been implemented in the Advanced Design System (ADS) software package using a multilayered neural network. The neural network model was trained to the drain current, gate-source capacitance and gate-drain capacitance characteristics of an Angelov model previously developed. Training was done using the back-propagation algorithm, which achieved excellent results for the model. Linear and harmonic balance simulations were performed to compare the linear and power performance of the neural network model to the Angelov model, without any additional optimization after the initial training.
Bibliography:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ISSN:0192-6225