Application of neural networks to spurious mode modeling in the design of multiplexing networks

Author

Shuqi Li ; Ying Wang ; Jesuthasan, Jude A. S. ; Ming Yu

Author_Institution

Dept. of Electr., Comput. & Software Eng., Univ. of Ontario Inst. of Technol., Oshawa, ON, Canada

fYear

2013

fDate

14-18 April 2013

Firstpage

1

Lastpage

4

Abstract

A method of modeling spurious mode for microwave multiplexer design is presented. It is proved that the channel filter can be simplified to have the minimum structural features, yet still capable of modeling the effect of spurious mode accurately, which facilitates the application of neural networks in the multiplexer design. Model simplification and neural model development are detailed. The method enables fast and accurate simulation of the spurious mode over wide frequency range. Excellent agreements are observed between the results obtained from EM simulations and from the developed models.

Keywords

channel bank filters; electromagnetic waves; microwave circuits; multiplexing equipment; neural chips; EM simulations; channel filter; frequency range; microwave multiplexer design; minimum structural features; model simplification; multiplexing networks design; neural model development; neural networks; spurious mode modeling; spurious mode simulation; Artificial neural networks; Band-pass filters; Data models; Integrated circuit modeling; Microwave filters; Multiplexing; Resonator filters; Computer aided design; multiplexer design; multiplexing network; neural networks; spurious mode;

fLanguage

English

Publisher

ieee

Conference_Titel

Wireless Symposium (IWS), 2013 IEEE International

Conference_Location

Beijing

Type

conf

DOI

10.1109/IEEE-IWS.2013.6616749

Filename

6616749