Determination of Complex Permittivity of Low-Loss Materials From Reference-Plane Invariant Transmission/Reflection Measurements

In this paper, a novel method is proposed to determine unique and stable relative complex permittivity of low-loss materials from the transmission/reflection (TR) measurements, though the positions of the materials under test are unknown. The proposed method systematically combines artificial neural network (ANN) models with the Nicolson-Ross-Weir (NRW) method. Firstly, an ANN model is established to estimate the position of the sample in the transmission line. Secondly, the NRW method is adapted to estimate the relative complex permittivity, and the unique solutions are selected with the help of dielectric properties. Thirdly, the noniterative technique in the literature is adapted to avoid the instabilities in the relative complex permittivity. Finally, the S-parameters of the amplitude-only or transmission-only measurements are used to accurately determine the relative complex permittivity. The proposed method is experimentally validated by two samples. The simulated data are used to further validate the accuracy of the proposed method. The results of the proposed method based on the transmission-only measurements are the most accurate.