Noise wave modeling of microwave transistors based on neural networks

被引：14

作者：

Markoviç, V ^{[1
]}

Pronic, O ^{[1
]}

Marinkovic, Z ^{[1
]}

机构：

[1] Univ Nis, Fac Elect Engn, Nish, Serbia

来源：

MICROWAVE AND OPTICAL TECHNOLOGY LETTERS | 2004年 / 41卷 / 04期

关键词：

microwave transistor; noise modeling; wave approach; neural networks;

D O I：

10.1002/mop.20120

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

The noise modeling of microwave FETs based on the noise-wave representation of a transistor-intrinsic circuit is considered. Frequency-dependent noise-wave temperatures are introduced as empirical model parameters and modeled using neural networks. In this way, online optimization in a circuit simulator is shifted to offline training of neural networks. An example of transistor-noise modeling for one specified component is shown. (C) 2004 Wiley Periodicals, Inc.

引用

页码：294 / 297

页数：4

共 9 条

[1]

[Anonymous], 1994, NEURAL NETWORKS

[2]

DOBROWOLSKI JA, 1991, INTRO COMPUTER METHO

[3]

*EESOF INC, 1990, TOUCHST LIBR US MAN

[4] WAVE APPROACH TO NOISE PROPERTIES OF LINEAR MICROWAVE DEVICES [J].