MACHINE LEARNING FOR SIGNAL MODELING OF GAN HEMTS

Abstract

High power and high frequency systems have been designed with better performance through the use of semiconductor devices called Gallium Nitride High Electron Mobility Transistors (GaN HEMTs)[1][2]. Hence, GaN HEMTs are used generally because they have large advantages over traditional semiconductors from Si and GaAs, especially in the areas of Radio Frequency (RF) and microwave applications[3]. As GaN-based devices have high thermal conductivity, a wide band gap, and can work at high voltages and temperatures, they are more preferable than the Si-based devices. GaN HEMTs have these characteristics, which make them very suitable for applications where high power and efficiency are required[4]. Hence, all this fact makes GaN HEMTs so common and used by so many engineers in the modern engineering communication systems (radar systems, satellite communications, high-performance amplifiers, wireless communication networks)...