A Power Amplifier Modeling Method For Digital Predistortion System In Noise Background

With the diversification of wireless communication application scenarios,the requirements for communication quality in different scenarios are getting higher and higher.Radio frequency power amplifiers are widely used in wireless communication systems,but they have nonlinear operating characteristics,so the quality of communication signals in the system will be reduced.In order to improve the nonlinear effects of power amplifiers on signals,many linearization technologies have been proposed.Among them,digital predistortion technology has great development potential in the industry due to its simple implementation principle and good nonlinear improvement effect.The performance effects of digital predistortion techniques are directly related to the accuracy of power amplifier modeling.Because the construction of the power amplifier model needs to use the input and output signals of the power amplifier.Therefore,in order to obtain a high-precision power amplifier model,it is necessary to ensure the quality of the signal involved in the calculation.In practical scenarios,the noise in the channel will interfere with the output signal of the power amplifier,thereby reducing the accuracy of the power amplifier modeling and further deteriorating the predistortion performance.Thesis will analyze the influence of noise on the power amplifier modeling accuracy,and give a power amplifier modeling method for the digital predistortion system in this scenario.The research is as follows:First,study the influence of power amplifier modeling on digital predistortion technology,focusing on the problem that noise affects the accuracy of power amplifier modeling.First of all,the influence of channel noise on the power amplifier modeling accuracy under different signal-to-noise ratios is compared,and the requirements for filtering out noise are determined.Then,according to the requirement of filtering out noise,a modeling method of power amplifier is designed,and the simulation results show that the method can improve the modeling accuracy of power amplifier.Second,the system implementation is carried out according to the requirements of the digital predistortion system in the noise scene.The operation of the modules and the design principle are analyzed in detail,and the module construction is completed in the FPGA(Field Programmable Gate Array).Among them,the algorithm used in the system is realized by the interaction between FPGA hardware and Micro Blaze soft core,including noise filtering,power amplifier parameter extraction,and predistortion lookup table making.Third,test the power amplifier modeling method.In two environments that only contain additive noise and both contain additive and multiplicative noises,the influence of the power amplifier modeling method designed in this thesis on the power amplifier modeling accuracy and the improvement effect on the system predistortion performance are analyzed.The test results show that compared with the power amplifier model constructed without filtering noise,in the environment with only additive noise,the power amplifier modeling accuracy can reach-68 dB,and the accuracy is improved by about 28 dB;In the environment of multiplicative noise,the power amplifier modeling accuracy can reach-54 dB,and the power amplifier modeling accuracy is improved by about 9dB.In these two environments,the ACLR(Adjacent Channel Leakage Ratio)value of the system output signal is improved by about 20 dB and 9dB,respectively.The power amplifier modeling method designed in thesis can effectively improve the accuracy and improve the predistortion effect.