Study and simulation of RF power amplifier behavioral model based on RBF Neural Network

Modeling power amplifier is a key step for designing power amplifying system and predistortion system. Whether nonlinearity and memory effects of power amplifier can be modeled correctly or not, has an important impact on system simulation performance. This paper presented and analyzed the Radial Basis Functions Neural Network (RBFNN), Utilizing input and output data extracted from Freescale semiconductor transistor MRF6S21140 model and designed circuit in ADS circumstance, simulate two kinds of Back Propagation Neural Network(BPNN) models and RBFNN models, and compute the error. Simulation results show that the proposed RBFNN behavioral model has less modeling error, and output waveform of the model can be more close to real waveform. The model can be set up in ADS and can be applied for system-level simulation, system simulation performance can be more close to real system, and has a significant sense for designing real system.