Finite-difference frequency-domain algorithm for modeling guided-wave properties of substrate integrated waveguide

• Published in: IEEE transactions on microwave theory and techniques Vol. 51; no. 11; pp. 2221 - 2227
• Main Authors: Feng Xu, Yulin Zhang, Hong, Wei, Wu, Ke, Tie Jun Cui
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2003; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Ceramics; Eigenvalues; Eigenvalues and eigenfunctions; Finite difference method; Finite difference methods; Mathematical analysis; Mathematical models; Matrix converters; Microwave integrated circuits; Microwaves; Nonhomogeneous media; Periodic structures; Printed circuit boards; Printed circuits; Rectangular waveguides; Studies; Transmission line matrix methods; Wave attenuation; Waveguides
• ISSN: 0018-9480; 1557-9670
• DOI: 10.1109/TMTT.2003.818935

In multilayer microwave integrated circuits such as low-temperature co-fired ceramics or multilayered printed circuit boards, waveguide-like structures can be fabricated by using periodic metallic via-holes referred to as substrate integrated waveguide (SIW). Such SIW structures can largely preserve the advantages of conventional rectangular waveguides such as high-Q factor and high power capacity. However, they are subject to leakage due to periodic gaps, which potentially results in wave attenuation. Therefore, such a guided-wave modeling problem becomes a very complicated complex eigenvalue problem. Since the SIW are bilaterally unbounded, absorbing boundary conditions should be deployed in numerical algorithms. This often leads to a difficult complex root-extracting problem of a transcend equation. In this paper, we present a novel finite-difference frequency-domain algorithm with a perfectly matched layer and Floquet's theorem for the analysis of SIW guided-wave problems. In this scheme, the problem is converted into a generalized matrix eigenvalue problem and finally transformed to a standard matrix eigenvalue problem that can be solved with efficient subroutines available. This approach has been validated by experiment.