An adaptive predistorter for high power amplifiers

In this paper we describe the performance of a new adaptive HPA (high power amplifier) predistorter which utilizes indirect learning architecture. This approach eliminates the necessity of first developing a fixed model of the HPA inverse amplitude response and direct phase response, since such responses are determined adaptively. The performance of the proposed predistorter is demonstrated via the reduction of spectral spreading, comparison of received 64-QAM signal constellations with and without distortion, and the ability of the predistorter to adapt to perturbations of the HPA characteristics.