Accurate EC-ANN modeling for a RF-MEMS extended tuning range varactor
- Resource Type
- Conference
- Authors
- Wang, Jie; Sun, Lingling; Liang, Yaping
- Source
- 2010 Asia Pacific Conference on Postgraduate Research in Microelectronics and Electronics (PrimeAsia) Microelectronics and Electronics (PrimeAsia), 2010 Asia Pacific Conference on Postgraduate Research in. :360-363 Sep, 2010
- Subject
- Components, Circuits, Devices and Systems
Bioengineering
Engineered Materials, Dielectrics and Plasmas
Fields, Waves and Electromagnetics
Signal Processing and Analysis
Integrated circuit modeling
Solid modeling
Varactors
Computational modeling
Micromechanical devices
Artificial neural networks
Training
Radio frequency micro-electro mechanical systems (RF-MEMS)
Equivalent circuit trained artificial neural network (EC-ANN)
Finite element methods (FEMs)
- Language
- ISSN
- 2159-2144
2159-2160
A novel accurate and efficient modeling method based on Equivalent Circuit trained Artificial Neural Network (EC-ANN) technique is developed for a RF-MEMS extended tuning range varactor. The parameters are extracted directly from the equivalent circuit model and used as training and testing sets for the ANN. Experiments show that the proposed approach can be used to fast and accurately model the RF characteristics of the RF-MEMS varactor. The results can agree with the EC-ANN predictions and the Ansoft HFSS simulations. To extend the capabilities of the proposed methodology, the developed EC-ANN modeling technique is used for design, simulation and optimization of the MEMS circuits.