Gallium nitride high electron mobility transistors (GaN HEMTs) boast exceptional properties, including high breakdown voltage, high power density, and the ability to operate at high frequencies. These features render them highly suitable for a wide range of applications, spanning wireless communications, RF power devices, and aerospace technology. The conventional design approach for GaN HEMTs hinges on TCAD simulation and parameter scanning techniques. To achieve specific cutoff frequency and maximum oscillation frequency targets, designers must conduct repeated simulations within a multi-dimensional parameter space. This process not only extends design cycles but also escalates computational expenses. Consequently, the development of swift and intelligent inverse design methodologies, tailored to meet specific performance goals, assumes paramount importance.
