Accurate and Low Complexity Polynomial and Neural Network Models for PA Digital Pre-distortion

In this paper, new polynomial and neural network models for power amplifier digital pre-distortion are introduced. The motivation behind the suggested models is having low complexity models that maintain good error performances. Also, this paper discusses the comparison between polynomial and neural network models in terms of model complexity and error performance before and after applying a compressed sensing algorithm. Furthermore, the proposed neural network model is a low complexity model that does not require a compressed sensing algorithm to reduce its number of model parameters, yet the proposed neural network model achieves good NMSE(dB) error performance.