Search

This paper proposes a scalable architecture for predicting antenna performance using various datainformed machine learning (DIML) methods. By utilizing the computation power of graphics processing units (GPUs), the architecture takes advantage of hardware (HW) acceleration from the beginning of electromagnetic (EM) full-wave simulation to the final machine learning (ML) validation. A total of 49152 full-wave simulations of a classical microwave patch antenna forms the ML dataset. The dataset contains the performance of patch antenna on six commonly used materials and two standard thicknesses in a wide frequency range from 0.1 to 20 GHz. A total of 13 base ML models are stacked and ensembled in a tabular workflow with performance as 0.970 and 0.933 F1 scores for two classification models, as well as 0.912 and 0.819 R2 scores for two regression models. Moreover, an image-based workflow is proposed. The image-based workflow yields the 0.823 R2 score, indicating a near real-time prediction for all S11 values from 0.1 to 20 GHz. The proposed architecture requires neither the fine-tuned hyperparameters in the ML-assisted optimization (MLAO) model for specified antenna design nor the pre-knowledge required in the physics-informed models. The fully automated process with data collection and the customized ML pipeline provides the architecture with robust scalability in future work where more antenna types, materials, and performance requirements can be involved. Also, it could be wrapped as a pre-trained ML model as a reference for other antenna designs. [ABSTRACT FROM AUTHOR]

This paper investigates the numerical and experimental analysis of a low-cost and dual-band microstrip patch antenna for the fifth generation (5G) mobile communications. The numerical analysis of the proposed antenna is performed using the computational electromagnetic simulator (CEMS) software which is based on the finite-difference time-domain (FDTD) and CST software which is based on the finite integration technique (FIT). The performance of the proposed antenna designed and fabricated on a low-cost FR-4 substrate is verified with the simulated and measured results. The antenna operates at dual frequency bands which are 24 and 28 GHz. The antenna maximum gain values are 3.20 dBi and 3.99 dBi in the x-y plane at 24 and 28 GHz, respectively. The proposed antenna provides almost omni-directional patterns suitable for 5G mobile communication devices. [ABSTRACT FROM AUTHOR]

A compact size arrow shaped patch in a rectangular slot antenna is designed for 5G communications in the lower 3 to 6 GHz band. The antenna element is fed through a coplanar waveguide with partial ground plane for better impedance matching with 50 Ohms across the entire band. The maximum gain of a single element is 3.8 dB at 3.7 GHz, while for linear arrays of 5 and 15 elements with uniform excitation the maximum gains are 10.9 dB and 16 dB, respectively. The 5 and 15 elements arrays provide scanning range with no significant degradation of the main beam up to 30˚ and 45˚, respectively. The properties of this antenna element makes it suitable for 5G wireless mobile devices and miniaturized base stations antenna arrays. [ABSTRACT FROM AUTHOR]