Metamaterial-based multiband 3D MIMO antenna systems design for 5G IoT applications with machine learning validation

Abstract

The rapid advancement of wireless communication technologies, particularly with the introduction of 5G and Internet of Things (IoT) applications, necessitates sophisticated antenna systems that support high data rates, improved connectivity, and a variety of functionalities. As the number of IoT-connected devices continues to rise globally, there is a pressing need for a wireless communication system capable of connecting numerous IoT devices via a single feedline antenna system. Such an antenna should facilitate both short-range and long-range communication by covering microwave and millimetre wave frequency bands with a single input. Multiple Input Multiple Output (MIMO) antennas are widely recognized for enhancing wireless communication stability and minimizing multipath fading loss. To further enhance the functionality of compact MIMO antennas, the integration of metamaterials (MTM) holds significant promise. This thesis focuses on the development and analysis of a metamaterial-integrated multiband and wideband MIMO antenna system optimized for 5G-IoT applications and designed for microwave and millimetre wave frequencies with 360-degree coverage and verify the antenna outputs using Machine Learning (ML) verification technique. By leveraging the unique properties of metamaterials, the proposed antenna system achieved superior performance in bandwidth, gain, and radiation efficiency across multiple frequency bands. The ML models, trained with a comprehensive dataset derived from simulated results, provide accurate predictions of critical parameters such as bandwidth, efficiency, and realized gain. Initially, a super wideband (SWB) 4-port MIMO antenna is developed incorporating metamaterials (MTM) operating within the 1.9 GHz to 20 GHz microwave spectrum. The integration of a metamaterial-based Metasurface improved efficiency from 77% to 89% and increased the realized gain from 4.5 dBi to 8 dBi, while isolation at the perimeter improved from 20 dB to 25.5 dB. Additionally, another 4-port MIMO antenna is developed with MTM to cover both microwave (3.5 GHz and 5.2 GHz) and millimetre wave bands (28 GHz). Metamaterials enhanced isolation levels from 20 dB to 24 dB in the microwave range and from 26 dB to 32 dB in the millimetre wave range, with maximum realized gains of 5.3 dBi and 9.3 dBi, respectively. To improve channel capacity and accommodate diverse IoT applications, developed a 3D 6-port MIMO antenna providing 360-degree radiation coverage for both microwave (2-7 GHz and 13-17.5 GHz) and millimetre wave (25-39 GHz) frequency bands, achieved a maximum realized gain of 9.9 dBi. Expanding on this, a 24-element 3D 12-port MIMO antenna integrated with MTM is developed, offering 360-degree radiation patterns in both microwave (2.5-8.4 GHz and 13.4-16.5 GHz) and millimetre wave (26-48 GHz) bands, with a maximum realized gain of 11.5 dBi. All antennas are simulated using CST simulation software, fabricated with Rogers RT-5880, and measured. Moreover, a machine learning verification technique was applied to validate antenna performance metrics, including gain, bandwidth, and efficiency, achieving over 90% accuracy and ensuring the reliability of the proposed antenna designs. The proposed 3D MIMO antennas, with their microwave and millimetre wave 360-degree coverage, demonstrate significant potential for future 5G wireless communication and IoT applications.