Design and Optimization of AI-Driven Terahertz Antennas for 5G and 6G Communication Systems

Abstract

This paper proposes an AI-driven multi-parameter optimization framework for terahertz antennas intended for 5G and 6G communication systems. The proposed methodology combines electromagnetic simulation with artificial intelligence and machine-learning-based prediction to establish the relationship between antenna design parameters and key performance indicators. Antenna geometrical parameters are systematically varied during the design-space exploration stage, while parameters such as resonant frequency, impedance bandwidth, reflection coefficient, gain, radiation efficiency, and radiation characteristics are considered as optimization objectives. A trained AI model is subsequently employed as a surrogate predictor to estimate antenna performance and identify promising design configurations before final electromagnetic validation. This approach reduces the dependency on repeated computationally intensive simulations and enables efficient multi-objective optimization. The proposed framework provides a systematic pathway toward intelligent THz antenna design and can be extended to adaptive and application-specific antenna optimization for future 5G and 6G communication platforms. Keywords— Terahertz antenna, 5G, 6G, artificial intelligence, machine learning, multi-parameter optimization, antenna design, electromagnetic optimization, surrogate modeling, wireless communication.