Electrothermal Design of Broadband Dual-Polarized Cross-Dipole Heatsink Antenna Using Multiphysics Machine Learning-Assisted Optimization

An all-metal dual-polarized cross-dipole antenna with fractal structures is designed to improve the electrothermal performance using a multiphysics machine learning-assisted optimization method. The internal simple heatsinks of the radiator, together with the external fractal heatsink structures, enhance the heat dissipation capability without compromising the electromagnetic (EM) performance. The fractal structure of the radiator is also used to improve the EM performance. The antenna prototype is fabricated and measured with a broad impedance bandwidth of 3.3–5.0 GHz, isolation higher than 27 dB, realized gain of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$6.5 \pm 1.4$</tex-math></inline-formula> dBi, and cross-polarization discrimination (XPD) greater than 20dB. Thermal simulation results show that compared with the conventional cross-dipole antenna without heatsink structures, the maximum temperature of the proposed antenna is reduced by 31.6%. Within the operating frequencies, a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$1 \times 3$</tex-math></inline-formula> subarray formed by the proposed antenna elements also shows excellent electrothermal performance. The simulated and measured results demonstrate that the proposed antenna and subarray are suitable for 5G New Radio frequency bands n77/n78/n79.