Design of a Multi-Band MIMO Microstrip Antenna Using Artificial Intelligence for 5G MM-Wave Applications.

This article presents the design of a multiband microstrip MIMO antenna for millimeter-wave (mm-Wave) applications, which are crucial for 5G networks. The antenna, designed using an Artificial Neural Networks (ANN) approach, specifically a multilayer perceptron, consists of four microstrip elements on a Rogers RT/Duroid 5880 dielectric substrate. The dimensions of the antenna are 31 × 22 × 0.79 mm². Operating at frequencies of 24, 31, and 36 GHz, the antenna is suitable for satellite communications, high-resolution radar, and point-to-point communications. The dimensions of the slots are determined using the ANN technique. A prototype of this MIMO antenna, featuring neutralization lines, has been simulated and tested, demonstrating mutual coupling of less than -24.8 dB between ports. The antenna offers an efficiency of 94.3%, gains of 7.63 dBi, 7.61 dBi, and 6.25 dBi at different frequencies, and a suitable radiation pattern. The Envelope Correlation Coefficient (ECC) is below 0.006, and the maximum Diversity Gain (DG) is approximately 9.98 dB. The Total Active Reflection Coefficient (TARC) is stable, meeting the requirements for various wireless MIMO applications. [ABSTRACT FROM AUTHOR]