NONLINEAR BEHAVIORAL OF GaN DOHERTY POWER AMPLIFIERS USING NEURAL MODELING

作者：Liu Haiwen^*

来源：Microwave and Optical Technology Letters, 2010, 52(2): 307-309.

DOI：10.1002/mop.24924

• GaN Doherty amplifier
• nonlinear behavioral modeling
• radial-basis function neural network (RBFNN)

A radial-basis function neural network (RBFNN) approach is proposed for predicting the nonlinear behaviors of gallium nitride (GaN) Doherty amplifier. Sampled input and output data from a designed GaN Doherty amplifier were used to train and test the proposed RBFNN model. Comparison of amplitude modulation to amplitude modulation (AM/AM), amplitude Modulation to phase modulation (AM/PM), gain, power added efficiency (PAE) and output power (Pout) curves among the RBFNN method, circuit simulation and measurement are given. The results indicate that the proposed RBFNN model can reproduce the nonlinear characteristics of the designed GaN Doherty amplifier accurately.

• 出版日期2010-2
• 单位华东交通大学