Multiband THz MIMO antenna with regression machine learning techniques for isolation prediction in IoT applications

Area of Science:

• Electrical Engineering
• Wireless Communication
• Antenna Design

Background:

• The Internet of Things (IoT) and wireless communication demand higher data rates and connectivity.
• Terahertz (THz) frequencies offer potential for next-generation wireless systems but require advanced antenna solutions.
• Existing MIMO antenna technologies face challenges in meeting the stringent requirements of THz applications.

Purpose of the Study:

• To develop and evaluate a novel Multiple-Input Multiple-Output (MIMO) antenna for Terahertz (THz) frequency bands.
• To enhance antenna efficiency and isolation using machine learning techniques.
• To assess the antenna's suitability for 6G wireless communication and Internet of Things (IoT) applications.

Main Methods:

• Antenna performance was assessed using simulation (CST) and RLC equivalent circuit models (ADS).
• The proposed MIMO antenna design operates at 6.51 THz, 7.48 THz, and 8.46 THz.
• Machine learning algorithms were employed to predict antenna isolation, with Gradient Boosting Regression showing superior performance.

Main Results:

• The antenna achieves compact dimensions (160 × 75 μm²) with high gain (13.53 dBi) and efficiency (>96.5%).
• Excellent isolation was recorded, exceeding -32 dB, -44 dB, and -45 dB across operating bands.
• The Gradient Boosting Regression model achieved over 98% accuracy in predicting isolation with low error rates (MAE 4.94%, MSE 6.60%, RMSE 4.13%).

Conclusions:

• The designed MIMO antenna is highly efficient and performs exceptionally well in the THz band.
• Machine learning significantly enhances the prediction accuracy of antenna isolation.
• This antenna design shows great promise for 6G wireless communication and IoT due to its performance characteristics.