A Novel Design of Non-Uniform Reflectarrays with Symbolic Regression and its Realization using 3-D Printer

Mahouti P., Güneş F., Belen M. A., Çalışkan A.

APPLIED COMPUTATIONAL ELECTROMAGNETICS SOCIETY JOURNAL, vol.34, no.2, pp.280-285, 2019 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 34 Issue: 2
  • Publication Date: 2019
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.280-285
  • Keywords: 3D printer, antenna design, microstrip reflectarrays, reflection phase modelling, symbolic regression, ANTENNA
  • Yıldız Technical University Affiliated: Yes


Herein, a novel design of an X-band Non-Uniform Reflectarray Antenna (NURA) with Symbolic Regression SR and its fabrication using 3D Printer technology are presented. A NURA is consist of simply a grounded dielectric layer with the variable thickness. Firstly, SR is employed to obtain with a great accuracy for the reflection phase characteristics of a grounded dielectric layer in the analytical form within the continuous domain of 1 <=epsilon(r)<= 6 and thickness 0.1 <= h <= 3 mm for the X-band. For this purpose, SR is trained and validated by the 3D CST Microwave Studio data. Then, for the design purpose of NURA, a special fine reflection calibration characteristic is built up again by SR with the sufficient reflection phase range of the 3D Printer's material at the operation frequency 10 GHz. In the third step, the designed NURA is then prototyped by using 3D printer technology where the material can be easily shaped and create unit cell at printing accuracy of 0.1 mm per layer. Thus, by this mean the prototyping cost of non-uniform Reflectarray design can be reduced drastically both in means of time and ease of manufacturing. In the final step, mismatching and radiation properties of the prototyped NURA are measured.