Research And Implementation Of Digital Predistortion Algorithm Based On Memory Polynomial Power Amplifier Model

Power amplifier is an important device in wireless signal transmission system,but it has nonlinear and memory characteristics,which will lead to the expansion of the transmission signal frequency band.It will make the signal distortion in the transmission process,and affect the receiving.Therefore,solving the non-linearity of power amplifiers is a hotspot in wireless communication transmission system.Digital predistortion technology has become the research focus because of its digitization,strong adaptability and high integration.Power amplifiers are divided into two categories according to whether they have memory effect.They have different applications in practical projects.The digital predistortion algorithms for memoryless and memory amplifiers are studied in this paper,in order to solve the distortion effect caused by different types of amplifiers in different situations.The following research work has been done around the topic of designing digital predistorter,accroding to the knowledge of non-linear model and adaptive filtering technology.A digital predistorter based on RASCAL algorithm is studied for memoryless amplifiers.The correction effect of this predistorter on the static distortion is analyzed by simulation.On this basis,combined with the theory of piecewise linear interpolation,a digital predistorter according to improved RASCAL algorithm is designed.By means of simulation,the improved digital predistorter has a better linearization effect from the aspects of graphics,performance indicators and so on.The simulation results show that the improved RASCAL algorithm has faster convergence speed and smaller error than the traditional RASCAL algorithm.For memory amplifiers,a digital predistorter depended on memory polynomial model with indirect learning structure is studied.The amplitude difference correlation function method is introduced in detail,which solves the problem of circuit delay in hardware implementation.The validity of this method is verified by simulation.Moreover,the digital predistorter based on odd order memory polynomial model and the digital predistorter based on even and odd order memory polynomial model are simulated respectively.The correction effect of each on memory nonlinearity is analyzed from the aspects of graphics and performance indicators.The influence of even order model coefficients on the predistortion effect is compared in this paper as well.The simulation results show that compared with the odd order memory polynomial model,the correction ability of the wideband signal spectrum broadening is improved by more than 3dB.The designed digital predistorter depended on the memory polynomial model is implemented on FPGA hardware platform.This paper introduces the hardware platform based on FPGA and the establishment of hardware experimental environment.From the logic level of the FPGA,the implementation scheme of the digital predistorter is introduced in detail.At last,the designed digital predistorter is verified by the combination of hardware and software.In the experiment,the QPSK modulation signal with center frequency of 2.7GHz and bandwidth of 12 MHz was pre-distorted.The model of power amplifier is ZRL 3500+.The output signal decreased about 33 percentage points compared with the unprocessed signal in EVM value,and the ACPR value decreased about 25 dB.