A new behavior model reconstruction algorithm for power amplifier

Abstract

5G wireless communication system sets a higher requirement on power dissipation of base station power amplifiers. In order to study the working characteristics of power amplifiers, it is essential to build a reasonable power amplifier behavior model. To reduce the complexity of power amplifier modeling and solution, this paper designs a greedy algorithm based on Dice criterion. Particle swarm optimization is adopted for optimization. In order to verify the advantages of the algorithm, this paper applies the adaptive sparse algorithm to the simplification of nonlinear memory polynomial power amplifier, and sets different simulation parameters and sparsity for practical testing. The simulation results show that the proposed algorithm can effectively improve the modeling accuracy and the convergence rate of the model coefficients.