AI-Augmented Geometry Evolution for a Wideband Slotted Radiator Antenna for 5G N78 Applications

Authors M. Jeyabharathi1 , D. Kumutha2 , G.V. Aswini2, P. Geetha3 , S. Anitha4
Affiliations

1Department of Electronics and Communication Engineering, Best Center, Sastra University, 110016 New Delhi, India

2Department of Electronics and Communication Engineering, AMC Engineering College, Bangalore, India

3Department of Electronics and Communication Engineering, Sathyabama Institute of Science and Technology, Chennai, Tamil Nadu, India

4Department of AIDS, Karpaga Vinayaga College of Engineering and Technonology, Chengalpattu, TN, India

Е-mail skvijaykumu@gmail.com
Issue Volume 18, Year 2026, Number 5
Dates Received 04 June 2026; revised manuscript received 19 October 2026; published online 28 October 2026
Citation M. Jeyabharathi, D. Kumutha, et al., J. Nano- Electron. Phys. 18 No 5, 05018 (2026)
DOI https://doi.org/10.21272/jnep.18(5).05018
PACS Number(s) 84.40.Ba
Keywords 5G antennas, n78 band, AI-assisted optimization, Bayesian optimization, ANN surrogate modeling, Geometry tuning, Wideband slotted radiator.
Annotation

With the objective to evolve the geometry of a compact wideband slotted radiator for sub-6 GHz 5G n78-band applications (3.0-4.2 GHz), this paper presents an AI-augmented methodology. To automatically improve slot shape, slot dimensions, feed parameters, and radiator contour, a hybrid workflow combining Bayesian Optimization (BO) and an Artificial Neural Network (ANN) surrogate model is presented. Throughout the 3.0-4.2 GHz range, the suggested three-stage evolution – from a simple monopole to a corner-truncated patch and ultimately to an AI-optimized polygonal slotted radiator – achieves continuous |S11|  − 10 dB bandwidth. Smooth impedance characteristics, enhanced current distribution, radiation efficiency of 78-86 %, and a realized gain of 5.0-6.1 dBi are all displayed by the optimized radiator. This method shows great promise for compact wideband radiators in 5G terminals by eliminating manual parameter sweeps and reducing design cycles.

List of References