High-performance quatrefoil-slotted THz MIMO antenna for 6G applications with regression-based machine learning validation.

With the emergence of every new generation of wireless communication, 6G will require ultra-high-speed, low-latency, and high-spectrum efficiency techniques, which would need sophisticated antenna designs working on the terahertz (THz) spectrum range. This article introduces a miniaturized 2-element MIMO microstrip antenna with the graphene radiation patch on a polyimide substrate and copper ground for the THz regime. The proposed antenna exhibits a resonant frequency at 4.78THz with the ultra-low return loss of - 52dB and has an operation bandwidth of 3.81-5.13THz, which achieves a wide bandwidth of 1.32THz in totality. The proposed antenna has a maximum gain of 11.97dB, high radiation efficiency reaching as high as about 90%, and a good isolation (- 40dB) is achieved between ports. The diversity gain (DG) and the envelope correlation coefficient (ECC) are 9.999 and 0.0000175, which confirms that the diversity performance is a good candidate for future-generation high data rate systems. An equivalent RLC circuit is proposed and verified by full-wave simulations in order to describe the impedance behavior of the antenna accurately. Moreover, ML regression algorithms such as Extra Trees Regression are incorporated in the design flow, which efficiently reduces the evaluation time and allows multi-parameter optimization. Comparative studies also verify that the proposed antenna outperforms existing antennas in terms of bandwidth, gain, isolation, and flexibility and is therefore a potential candidate for 6G communication, biomedical imaging, and high-resolution sensing. The proposed antenna, which exhibits compact size, superior isolation, and high efficiency, yields tremendous potential for 6G applications at high Data Rate and presents a strong candidate to operate in the next generation of wireless communication devices.