AI-Driven Metasurface Assisted Graphene Based Reconfigurable Antenna For Terahertz Communication

Abstract: This paper presents the performance enhancement of graphene-based pattern reconfigurable antennas for terahertz (THz) applications. An AI-driven metasurface assisted graphene antenna with adaptive beam steering is proposed. Unlike conventional designs that uses bias voltage control for beam reconfiguration, this method integrates a metasurface layer to enhance electromagnetic wave manipulation in the frequency range 4 - 6 THz, improving gain 16 dBi, bandwidth, and directionality. Additionally, an AI-based control system dynamically adjusts the chemical potential of graphene elements in real time, enabling continuous 360° beam steering instead of fixed beam states. The system utilizes Random Forest machine learning algorithm to predict optimal bias voltages based on environmental conditions, ensuring real-time adaptation for enhanced signal strength and reduced interference. This next-generation design provides higher gain, broader coverage, and intelligent beam adaptation, revolutionizing high-speed THz wireless communication for future smart networks.

Keywords: Graphene-based antenna, Pattern Reconfigurable Antenna, Terahertz (THz) communication, Wireless communication systems.