Fostering RF Engineering Competencies Through AI-Integrated Project Based Learning

Abstract

Within the past few years, engineering education has been more and more oriented towards practical, real-world problem solving as well as theoretical education. This change is especially crucial in aspects like electromagnetic wave theory, which supports current 5G and new wireless networks. Nevertheless, Microwave Engineering and Antenna Design courses are commonly lecture-based, and students do not have the skills to solve more complex design problems. In order to close this gap, this work suggests a Project Based Learning (PBL) model of undergraduate RF training. Four patch antenna designs are designed, simulated and analyzed in microstrip with 28 GHz operations, and allow comparison of geometric variations between the designs. In this context, an AI-assisted optimization framework is applied whereby design parameters are produced algorithmically, tested against performance goals, and improved over time to arrive at optimal performance. The most important metrics used to analyze the performance of antennas are return loss (S11), gain, directivity, radiation efficiency, and VSWR. The learning is based on the CDIO (Conceive–Design–Implement– Operate) approach. Findings show enhanced problem-solving skills, design knowledge, and simulation skills and general performance improvement of around 32 percent over traditional approaches.