Modeling method of GaN HEMT small signal model based on support vector machine

Abstract

The invention discloses a modeling method of a GaN HEMT small signal model based on a support vector machine. The modeling method comprises prediction of behavior characteristics under different biasstates and working frequency conditions. According to the transistor behavior characteristic modeling method, the support vector machine technology is adopted, an error correction model is added on the basis of an existing equivalent circuit model according to the input and output change rule of the transistor, the accuracy of the output behavior characteristics of the GaN-based transistor is further improved, and high-precision transistor model establishment is achieved. An existing model established based on an equivalent circuit method is improved, and modeling efficiency and modeling precision are improved.

Description

technical field [0001] The invention relates to the field of modeling of microelectronic devices, in particular to a modeling method of a GaN transistor small-signal behavior model based on a support vector machine. Background technique [0002] Gallium nitride (GaN) is a semiconductor with a direct energy gap. It is an extremely stable compound with a high degree of ionization, which is the highest among III-V compounds. The research and application of GaN materials is the frontier and hot spot of global semiconductor research at present. It is a new semiconductor material for the development of microelectronic devices and optoelectronic devices. Together with semiconductor materials such as SIC and diamond, it is known as the successor to the first generation of Ge, Si Semiconductor materials, second-generation GaAs, and the third-generation semiconductor materials after InP compound semiconductor materials. The wide band gap semiconductor GaN material has excellent chara...

AI Technical Summary

Problems solved by technology

[0006] The present invention overcomes the deficiencies of the prior art, and proposes a small-signal behavior model modeling method for GaN HEMT devices based on support vector machines, which solves the problem of low modeling accuracy of the existing GaN HEMT device behavior models. Based on the equivalent circuit model, an error correction model (Error model) is added to make the output behavior characteristics of GaN transistors more accurate, and a model that can accurately predict the nonlinear behavior of GaN HEMT devices is established.

Images