Design and development of miniaturized MIMO antenna using parasitic elements and Machine learning (ML) technique for lower sub 6 GHz 5G applications

The proposed work is devoted to the development of a parasitic elements-based MIMO antenna with low mutual coupling using Machine Learning (ML) technique for lower sub 6 GHz 5G applications. This design solves and analyzes the complexity of the optimum position determination of parasitic elements and ground plane dimensions of composite MIMO antenna structure using ML algorithm. The development process for ML modeling technique is introduced and discussed. The suggested MIMO antenna using parasitic elements is implemented on a 1.6 mm thick FR-4 substrate of 50 mm x 30 mm. The proposed MIMO antenna shows an impedance bandwidth of 2000 MHz (3-5 GHz) with a fractional bandwidth of 50%. The suggested MIMO structure covers the lower sub-6-GHz 5G NR frequency bands n77 (3300-4200 MHz), n78 (3300-3800 MHz), and n79 (4400-5000 MHz). The performance of the MIMO structure in terms of mutual coupling, return loss and Envelope Correlation Coefficient (ECC) etc. are analyzed using the approach of Finite Element Method. Finally, the prototype of the proposed MIMO antenna is fabricated and the experimental results show good agreement with the simulated results. The proposed antenna offers radiation efficiency > 80%, isolation > 28 dB, ECC < 0.0000495, DG > 9.999 dB, CCL < 0.017 bits/s/Hz.