The Research On Digital Predistorter And Optimization Algorithm For 5G Power Amplifiers

With the development of wireless communication systems,the fifth-generation mobile communication system(5G)has become an important carrier of an intelligent society formed by smart homes,smart medical treatment,and smart cities.As one of the key components of 5G communication systems,power amplifiers directly determine the performance of wireless communication systems,especially the inherent non-linearity of power amplifiers,which will seriously degrade the communication quality.And a modulation signal with a high degree of modulation is used in the 5G communication system,which places stringent requirements on the linearity of the power amplifier.In addition,in order to meet the communication rate,the bandwidth of 5G systems has become wider,resulting in more complex memory effect produced by power amplifiers.So how to deal with the nonlinearity and memory effect of 5G power amplifier is the important and difficult problem in the current wireless communication research.Digital predistortion(DPD)technology has the advantages of strong linearization performance,reconfigurability,and self-adaptation.It is a very effective method to solve the nonlinear problem of power amplifiers.Therefore,this paper builds the model of 5G power amplifiers based on DPD technology,and study optimization on the model.The innovations of this paper are mainly reflected in: 1)The optimization algorithm of the power amplifier model is proposed by using the Attention mechanism in natural language processing;2)The 5G power amplifier behavior model is proposed by combining the Chebyshev polynomial and the LSTM network.Firstly,we analyze the nonlinearity of the power amplifier,and introduce the characteristics and application scenarios of the existing linearization technology.With the enhancement of the nonlinearity of the power amplifier,the complexity of the model describing the power amplifier is also getting higher,which is not conducive to the application of digital predistortion technology and cost control.Therefore,in this paper,an approach using the attention mechanism is proposed,which can be used for reducing the complexity of the DPD model.The attention mechanism is employed to obtain the weighted correlation coefficient matrix of the memory effect.And the memory terms in the DPD model will be retained only if the contributions are high,which are evaluated by ensemble averaging over each diagonal of the weighted correlation coefficient matrix.To verify the applicability of the approach,a three-carrier wideband code-division multiple access signals(3CWCDMA)with a bandwidth of 15 MHz and a single carrier long-term evaluation signal(1C-LTE)with a bandwidth of 20 MHz are employed for testing two Doherty RF-PAs with an operating frequency of 460 and 1900 MHz.Moreover,the generalized memory polynomial(GMP)model is used to verify the effectiveness of the proposed approach.The simulation and experimental results show that the modeling accuracy and the DPD linearization performance of the DPD model with and without the memory term reduction are all almost the same,which indicates the effectiveness of the proposed approach.In addition,in order to describe the complex nonlinearity of 5G power amplifiers correctly,this paper design a Long Short Term Memory(LSTM)network model based on Chebyshev polynomial expansion using the deep learning technology.The model mainly includes a data processing layer,a Chebyshev polynomial extension layer,an LSTM network layer,and a fully connected layer.Among them,Chebyshev polynomial expansion can well fit the envelope characteristics of the power amplifier signal,and the LSTM network layer can well reflect the signal timing,which can more effectively describe the memory effect of the power amplifier and have great significance in improving the accuracy of the model.In order to verify the linearization effect of this model,a singlecarrier LTE signal with a bandwidth of 20 MHz and a 5GNR signal with a bandwidth of 100 MHz are used to test a class AB power amplifier with a working frequency of 28 GHz.Excluding the limit of the instrument's sampling bandwidth,the model proposed in this paper can improve the adjacent channel power ratio(ACPR)by up to 18 d B for a 28 GHz power amplifier which use a 5GNR signal with a 100 MHz bandwidth as the input signal.