An adaptive predistorter using modified neural networks combined with a fuzzy controller for nonlinear power amplifiers

In digital radio systems, high data transmission rates require the use of spectrally efficient linear modulation techniques; however, these techniques are generally sensitive to nonlinearity caused by the high-power amplifier (HPA) employed in transmitter systems. The nonlinearity of HPA is potentially responsible for spectral spreading, adjacent channel interference (ACT), and degradation of bit-error rates (BERs). This article proposes an adaptive predistortion scheme to compensate for the HPA's nonlinearity by combining adaptive structure-varying neural networks and a fuzzy controller. Simulations show that this predistortion scheme can very effectively prevent the warping of the signal constellations, thus reducing the system's BER and learning time. (C) 2003 Wiley Periodicals, Inc.