Power Amplifier Linearization Using Multi- Stage Digital Predistortion Based On Indirect Learning Architecture

摘要

Power amplifiers (PA) are one of the essential components of communication systems and are nonlinear in nature. The nonlinearity creates in band and out of band distortions. To linearize a PA the cost effective method is use of digital predistortion. In this paper we propose a multi stage algorithm for digital predistortion using QR Decomposition. The coefficients are estimated using Indirect Learning Approach (ILA). In multi stage ILA the predistortion is implemented in two or more stages as compared to the single stage implementation of the conventional ILA approach. The multistage predistorters can achieve the same performance or even better performance than single stage predistorter depending on the power amplifier with lower complexity. The complexity is measured by the number of coefficients required for the identification of the predistorter. The performance of the multistage ILA is evaluated in terms of improvement in spectral regrowth suppression when an OFDM signal is given as input signal. Wiener-Hammerstein model is used for PA modelling.