Low-Complexity Model Optimization Design And Implementation For Wideband Power Amplifier

Power Amplifier(PA)is a critical device that determines the performance of communication.To ensure the power required during.signal transmission,it usually works in the saturation region.This will lead to nonlinear distortion and memory effects in the power amplifier,reducing the efficiency of the power amplifier and causing spectral leakage.In order to compensate for the distortion of the power amplifier,Digital PreDistortion(DPD)has been widely used as a linearization method.In the existing power amplifier behavior model,a generalized memory polynomial(GMP)is generally used to characterize the nonlinear behavior of the power amplifier.The GMP model has high accuracy,but also has high complexity.Researchers have proposed a pruning algorithm represented by doubly orthogonal matching pursuit,which can obtain the GMP model with the minimum complexity.However,when DPD is actually applied,a look-up table.is generally used to construct a digital predistortion block for compensating the power amplifier.The minimum complexity of the GMP model does not represent the minimum complexity of the look-up table.Excessive lookup table complexity will increase hardware cost and system power consumption.To further reduce hardware resource costs and reduce the hardware implementation complexity of look-up table-based digital predistortion,this paper proposes a low-complexity digital predistorter design method based on target optimization.It achieves the effect of reducing the number of look-up tables.In this paper,the optimization design problem of predistorter based on lookup table is established.Under the condition of guaranteed digital predistortion performance,the basic principle of digital predistorter is constructed according to lookup table,and the relationship between predistortion model basis function and lookup table is deeply explored,and optimize the predistortion model with the goal of minimizing the number of lookup tables.The experimental results show that,compared with the existing low-complexity model optimization design algorithms,the proposed method can build a digital predistorter with a low resource occupancy rate based on the look-up table,and ensure its high-precision compensation for the nonlinear distortion of the power amplifier.