A neurocomputational model for estimating the triple-frequency of T-shaped patch antennas

Yigit E., Kayabasi A., Toktas A., Sabanci K.

MICROWAVE AND OPTICAL TECHNOLOGY LETTERS, vol.61, no.6, pp.1590-1597, 2019 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 61 Issue: 6
  • Publication Date: 2019
  • Doi Number: 10.1002/mop.31831
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.1590-1597
  • Bursa Uludag University Affiliated: No


This article deals with the analysis of T-shaped patch antennas (TPAs) that operates between 1 and 7 GHz at triple-band characteristics. A TPA is composed of three monopole structures so that it has triple-resonant frequency. Neurocomputational (NC) models eliminate the complex procedures for the analysis of patch antennas with irregular shapes. In this study, a NC model based on artificial neural network (ANN) is constructed for analyzing the triple-frequency of TPAs. One hundred TPAs with different electrical and geometric parameters are simulated with a full-wave electromagnetic simulator, and a data matrix is obtained for the training and testing the NC model. The model is trained through the simulated data vector of 80 TPAs and is tested with the remainders 20 TPAs and a fabricated TPA. Therefore, the computed results by the NC model which estimates simply and fast the operating triple-frequency of TPA agree well with the simulated and measured ones.