João is pursuing his PhD in Electrical Engineering, hosted by IT in Aveiro, under the supervision of José Pedro, Luís Nunes (both from IT/UA) and Nikolai Sobolev (UA). (see more)





João Gomes, a PhD student from the University of Aveiro (UA) and a research assistant with the IT in Aveiro has won a Graduate Student Fellowship in the amount of $6,000 sponsored by the largest scientific society in the microwave area, the IEEE Microwave Theory and Techniques Society - MTT-S, for his work “Nonlinear Modeling of GaN HEMTs for RF and Microwave Applications”.

A wireless communication system uses a power amplifier (PA) to increase the signal amplitude before transmitting it through the antenna. These power amplifiers are the components that most consume power in a base-station, which led the telecom industry and scientific community to invest into more energy-efficient PA designs. Furthermore, faced with the new 5G requirements, the complexity of PA systems for wireless communications has been quickly escalating with the increase in signal bandwidth and the required efficiency and linearity. In addition, the adoption of higher frequencies and monolithic implementations has ascribed an essential role to nonlinear models for PA design.

Unfortunately, the state-of-the-art models of RF power transistors are not sufficiently accurate to design a PA based only on computer-aided design simulations, especially for technologies such as high electron mobility transistors based on Gallium Nitride (GaN-HEMTs) that are quickly becoming a promising next-generation power semiconductor. The biggest challenge to obtain accurate nonlinear models is how to mimic the low-frequency dispersion behavior that are normally observed in these transistors. This insufficiency of the representative and predictive capabilities of modern models is notably revealed by the nonlinear capacitance models’ internal inconsistency between the charge and energy conservation principles.

The objective of this project was thus, to discover the fundamental sources of these inconsistencies and then develop new equivalent circuit compact models of GaN HEMTs that can produce better estimations of real high-frequency measured data while maintaining high computational efficiency.

In the 2020 MTT-S Graduate Student Fellowship Awards, 47 applications from 14 countries were received in the general category, and 10 applications from four nations were received in the medical area. Twelve graduate fellowships were awarded for 2020 in the general category and two in the medical applications domain.