Optimal sizing of generalized memory polynomial model structure based on Hill-Climbing heuristic

Abstract

Digital Predistortion (DPD) is used to compensate the nonlinearities and memory effects of the Power Amplifiers (PA). Generalized Memory Polynomial (GMP) is usually applied for PA as well as predistorter modeling. However there are very huge number of different GMP model structures. In this paper, we propose an algorithm based on Hill-Climbing algorithm to search for the GMP model structure which provides the best tradeoff between modeling accuracy and its complexity. The proposed method has been evaluated on real Doherty Power Amplifier. The results show that the proposed algorithm allows to decrease the searching time by a factor typically greater than 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sup> .