Adaptive Signal Separation for Dual-Input Doherty Power Amplifier

In this article, the generating mechanisms of nonlinear distortion of the dual-input Doherty power amplifier (DIDPA) are discussed, and based on this, an adaptive signal separation algorithm (ASSA) is proposed to improve the nonlinearity of DIDPA. The main idea of the ASSA is to minimize the mismatch between the signal separation functions and the intended operation, which is considered to be the leading cause of nonlinearity in DIDPA. The signal separation functions are used to determine the power ratio and phase difference between the main and the auxiliary amplifier. Moreover, in order to maximize the drain efficiency, an efficiency dictionary which is derived from the static measured results is employed during the ASSA. The performance of the ASSA is validated experimentally on a gallium nitride (GaN)-based DIDPA. The measured results show that with the help of the ASSA, the adjacent channel power ratio and normalized mean square error of output signal can be improved significantly without sacrificing efficiency. In order to further demonstrate the validity of the ASSA, the digital predistortion results of the DIDPAs with and without ASSA are compared with two kinds of behavioral models and two kinds of modulation signals.