Comparative analysis of power amplifiers' polynomial based models identification using RLS algorithm

This paper investigates the performance of RF power amplifiers' behavioral models in the context of the adaptive model coefficients' identification. The forward twin-nonlinear two-box (TNTB) model is compared to the memory polynomial and orthogonal memory polynomial models. The study, performed using measured data of a Doherty power amplifier prototype driven by multi-carrier signals, highlights the complexity reduction provided by the TNTB model in comparison with the two other models. The results show the superiority of the TNTB model in the context of adaptive parameter-estimation as it leads to better normalized mean squared error while requiring a substantially lower number of parameters. Furthermore, the TNTB model requires less parameters for its identification, and thus less power consumption for its estimation. This makes this model suitable for implementation in energy efficient green communication systems.