Robust formulations of the cauchy method suitable for microwave duplexers modeling

Daniele Traina, Giuseppe Macchiarella, Tapan K. Sarkar

Research output: Contribution to journalArticle

19 Scopus citations

Abstract

It is well known that a Vandermonde matrix generates an ill-conditioned system matrix when applied with finite numerical precision. This deficiency affects the Cauchy method by restricting its application to only lower order systems. This paper presents innovative, accurate, and robust formulations of the Cauchy method to rectify this limitation and make the Cauchy method suitable for the extraction of a high-order microwave duplexer polynomial model. The techniques employed are: the change of polynomial basis into the Krylov subspace and the precondition technique, both acting on the system matrix of the classic Cauchy method formulation. A novel formulation using the QR algorithm on the two characteristic functions of the duplexer and a suitable combination of the QR method and the precondition technique are then presented. Each of these procedures has been successfully verified by numerical application examples.

Original languageEnglish (US)
Pages (from-to)974-981
Number of pages8
JournalIEEE Transactions on Microwave Theory and Techniques
Volume55
Issue number5
DOIs
StatePublished - May 1 2007

    Fingerprint

Keywords

  • Cauchy method
  • Condition number
  • Microwave duplexer
  • Numerical analysis
  • Robust modeling

ASJC Scopus subject areas

  • Radiation
  • Condensed Matter Physics
  • Electrical and Electronic Engineering

Cite this