Behavioral Modeling and Digital Predistortion of RF Power Amplifiers

Modeling accurately the nonlinear behavior of radio frequency power amplifiers is a key point in the design and development of linearizer. In this paper we presented a predistorsion technique based on DSP circuit and the identification of continuous time model describing the power amplifier associated to quadrature modulator and demodulator circuits. The model is composed by two blocks, the first uses polynomial functions to model the nonlinear amplitude and phase distortion, the second one is based on a MIMO continuous filter to describe the PA dynamics. Parameters are determined using identification method based on output error technique and a pseudo random binary sequence as signal excitation. Using the previous solution, the non-linear inverse function is calculated and implemented in the DSP. Experimental results of digital predistorsion show an improvement of 10 dB of the ACPR for QPSK and pi/4 DQPSK modulated signals