COMMITTEE CHAIR: Dr. Lijun Qian
TITLE: AI-ENABLED DESIGN AND PROTOTYPING OF ANTENNAS FOR FUTURE WIRELESS SYSTEMS
ABSTRACT: The increasing demand for rapid and reliable antenna prototyping in emerging 5G/6G systems has intensified the need for computationally efficient electromagnetic (EM) design methodologies. Traditional full-wave simulations remain accurate but are computationally expensive, especially for broadband and inverse design tasks. In this PhD research, an AI-accelerated computational electromagnetics framework is proposed that unifies data-driven learning, physics-informed modeling, and forward-consistent inverse design for better design and fast prototyping of UWB patch and fork antennas. High-fidelity EM simulations are used to construct parametric datasets linking antenna geometries to their corresponding return-loss spectra, bandwidths, and resonant frequencies. Machine learning models including fully connected deep neural networks, LSTMs, and reservoir computing architectures are developed to predict antenna dimensions directly from performance targets, significantly accelerating early-stage UWB design exploration. The proposed framework delivers a unified, physics-aware, AI-accelerated pipeline for fast EM prototyping, enabling rapid inverse design, broadband validation, and geometry optimization of patch and UWB antennas. The approach bridges the gap between traditional computational electromagnetics and modern deep learning, offering a promising direction for next-generation antenna design automation and fast on-demand RF prototyping.
Keywords: Artificial Intelligence, Machine Learning, Antenna Design, Neural Network, Wireless Systems
Location Online:
Zoom Link:
https://pvpanther.zoom.us/j/94376892289?pwd=130k5RZpFIKEu2SbA5cl61iKzfKaTU.1
Meeting ID: 943 7689 2289
Passcode: 974092