SEARCH

SEARCH BY CITATION

Keywords:

  • dielectric constant;
  • loss factor;
  • numerical modeling;
  • MLP neural networks;
  • reflection coefficient

Abstract

A simple novel cavity-independent method of determination of dielectric properties of arbitrarily shaped materials is presented. Complex permittivity is reconstructed using a neural networking procedure matching the measured and FDTD-modeled frequency characteristics of the reflection coefficient. High accuracy and practical suitability are demonstrated through numerical testing and determination of dielectric properties of fresh and saline water at 915 MHz. © 2004 Wiley Periodicals, Inc. Microwave Opt Technol Lett 40: 183–188, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop.11323