Training Data Pre-Processing for Bias-Dependent Neural Models of Microwave Transistor Scattering Parameters
Authors: Z. Marinković, V. Marković
Keywords: Microwave transistors, scattering parameters, artificial neural networks
Abstract:
Frequency and bias dependence of scattering parameters of microwave transistors can be successfully modelled by artificial neural networks. But, sharp changes in the frequency dependence of angle of a scattering parameter may result in inappropriate modelling accuracy in the vicinity of the frequency when sharp change occurs. In this paper we are discussing pre-processing of the training data in order to make modelling accuracy better. The proposed approach is illustrated by a suitable example.
References:
[1] Q. J. ZHANG, K. C. GUPTA, Neural Networks for RF and Microwave Design, Artech House, 2000.
[2] F.GUNES, H.TORPI, F.GURGEN, Multidimensional signal-noise neural network model, Circuits, Devices and Systems, IEE Proceedings, Vol.145, Iss.2, Apr 1998, pp. 111-117.
[3] Z. MARINKOVIĆ, A. STOIĆ, V. MARKOVIĆ, O. PRONIĆ, ANNs in Bias-Dependent Modeling of S-parameters of Microwave FETs and HBTs, Microwave Review, No.1, Vol. 12, June 2006, pp. 21-30
[4] D. POZAR, Microwave Engineering, John Wiley & Sons, Inc., New York, 1998
[5] http://www.semiconductor.agilent.com.