Digital Predistortion Technology Based On Space Coupling

Power amplifiers are an important part of communication systems.There is a contradiction between the linearity and the efficiency of the power amplifier.When the amplifier is operating in a linear manner,the efficiency is not high.When operating in a non-linear region,non-constant envelope modulation results in in-band distortion and out-of-band spreading of the spectrum of the signal.In order to maintain the high efficiency and low distortion of the amplifier,it is necessary to use a linearization technique to compensate for the non-linearity of the amplifier.Predistortion technology is a promising linearization technology.This thesis studies the digital predistortion technology and proposes a scheme to reduce the predistortion feedback measurement channel for large-scale MIMO predistortion technology,and validates it in the 5G experimental platform of the State Key Laboratory of Mobile Communications of Southeast University.This thesis first analyzes the performance of nonlinear characteristics of power amplifiers theoretically.Then the memory effect of the amplifier is analyzed and the memory performance of the amplifier is measured.Afterwards we introduce the commonly used linearization technology and contrast the advantages and disadvantages of a variety of linearization technology.Subsequently,some problems in the predistortion system are analyzed.The predistortion architecture adopts an indirect learning structure.The link delay uses integer multiple estimation and fractional multiple estimation.The convergence algorithm of PA model parameters extraction is analyzed and polynomial modeling and neural network modeling are performed.The effective order of power amplifier modeling is determined by the measured data.The predistortion verification platform is built to verify the consequence of these two models.For the large hardware overhead of the traditional pre-distortion scheme of large-scale MIMO systems,a wireless pre-distortion scheme is proposed,which can effectively lower the pre-distortion feedback channel.However,the influence of the channel on the pre-distortion performance needs to be considered.Finally,we use the 5G experimental platform to test the channel characteristics of the 100 MHz signal bandwidth in the3.5GHz band.Two technical solutions to the impact of the channel on the pre-distortion data is presented,and the effectiveness of the solution is successfully verified on the 5G experimental platform.