Generative AI-Powered Platform for Antenna Design

Generative Artificial Intelligence (GAI) has permeated virtually every aspect of human endeavor, revolutionizing processes and decision-making across diverse fields. In telecommunications engineering, 3D design plays a pivotal role, particularly in areas such as antenna design and metasurface unit cell development. This paper introduces an innovative Large Language Model (LLM)-assisted framework for the autonomous generation of antenna design. Based on this framework, we implement a proof-of-concept GAI platform that integrates the ChatGPT API with Ansys Electronics' HFSS module, enabling the creation of an antenna lens through generative GAI-driven prompts. The framework incorporates two optimization loops: an initial creative design phase and a subsequent fine-tuning phase, allowing for both innovative exploration and precise optimization. To demonstrate the effectiveness of our approach, we present a dual-frequency slot antenna design operating at 5 GHz and 5.75 GHz, featuring a GAI-generated lens structure for enhanced performance. This novel LLM-assisted framework and its corresponding GAI platform demonstrate the potential to streamline and enhance the design process for telecommunication hardware components, opening new avenues for innovation in electromagnetic engineering while leveraging human expertise in the design validation and refinement stages.