Adaptive Digital Predistortion of RF Power Amplifiers Based on Memory Polynomial Model and Indirect Learning Architecture

Abstract

This paper presents approaches for modeling and simulating RF power amplifiers and increasing transmitter linearity using adaptive digital predistortion technique. The memory polynomial models, indirect learning architecture and coefficients update algorithms such as recursive prediction error method, Kalman filter and recursive least squares are considered. The simulation results in the DVB-T2 standard indicate that an adjacent channel leakage ratio correction of over 25 dB is obtained due to linearization.