A Novel Neuro-Space Mapping Technique Incorporating Self-heating Effect for High-Power Transistor Modeling

Accurate modeling of self-heating effect of high-power transistor is critical for reliable design of microwave circuit and system. In this paper, a novel neuro-space mapping (Neuro-SM) method incorporating self-heating effect is presented. By modifying the voltage and temperature relationships in the existing electro-thermal nonlinear model, the proposed Neuro-SM produces a new model exceeding the accuracy limit of the model. To accurately describe the self-heating effect, separate mappings for temperature and voltage at gate and drain are used as the mapping structure in the proposed method. The mappings combined with thermal sub-circuit including thermal resistance-capacitance parallel with thermal current are used to describe the self-heating effect. The validity and efficiency of the proposed Neuro-SM method incorporating self-heating effect are demonstrated through a modeling example of a high-power transistor used in cellular infrastructure market.