Design and Analysis of Four-Port MIMO System Optimization Methodology with Machine Learning Approaches of Validated Antenna Parameters

A four-element MIMO array antenna for the generation of 5G cellphones to operate at lower sub-6 GHz systems and which is optimized using the current design approach. For acceptable performance related to the optimum design parameters EM simulations are performed instead of traditional optimizations method due to time-consuming process for antenna design. Depending on the nonlinear relationship and complexity for design characteristics an effective method of deep learning (DL) is to determining optimum physical parameters. For designing of the MIMO antenna array this technique proposes resource efficient and time using DL approach. To reduce design space and generation of an effective dataset the technique applied is feature reduction method. To predict the S-parameters developing a novel dual-channel deep neural network. The designed antenna is optimized and achieved the − 10 dB impedance bandwidth of 40% (2.8–4.2 GHz), ECC is 0.018, TARC value is − 22 dB, isolation is − 35 dB, mean effective gain ratio is approaches to unity and CCL is 0.32 bps/Hz.