Research On 2-D Orthogonal Polynomial Digital Predistortion For Concurrent Dual-Band Power Amplifiers

Carrier aggregation is one of the key technologies in the next generation mobile communication system.In order to achieve this technology,it is necessary to use a specific structure of the power amplifier to support multi-band/multi-standard signal,thereby significantly reducing the cost of wireless transmitters and simplifying the wireless transmission system.However,dual-band transmissions exhibit very high peak-to-average power ratios,and when they are combined with the nonlinearity behavior of a PA,can generate in-band distortion,spectral regrowth and inter-modulation distortion,which lead to unwanted spectral emissions in the adjacent channels or bit error rate deterioration in the receiver.So,the linearization of the dual-band power amplifier has become one of the key technologies in wireless communication systems.Among the many power amplifier linearization techniques,the most widely used technique is the digital predistortion technique,it allows PAs to be operated at high-efficiency status without sacrificing linearity,and tends to be a fundamental component in current and upcoming wireless transmitters for its robustness,high accuracy and moderate implementation cost.In this paper,we proposed a 2D orthogonal polynomial predistortion model for complex Gauss signal,based on the advantages and disadvantages of the traditional 2D polynomial predistortion model and the modified Legendre polynomial predistortion model.This algorithm can effectively reduce the matrix condition number,improve the numerical stability,and cut down the extraction of predistortion coefficients,reduce the computational complexity.In the paper,we make a simulation by OFDM and WCDMA signals,and validate that when the nonlinear order is 9,the condition number of 2D orthogonal polynomial model compared with conventional 2D polynomial model were reduced by 1055 magnitude(OFDM signal)and 1023 magnitude(WCDMA signal),and compared with modified Legendre model were reduced by 1013 magnitude(OFDM signal)and 1014 magnitude(WCDMA signal).In the performance of ACPR,the value of 2D orthogonal polynomial model is 45dBc(OFDM signal)and 30dBc(WCDMA signal),the value of modified Legendre polynomial model and the conventional 2D polynomial model is approximately 25dBc(OFDM signal)and 20dBc(WCDMA signal).In the 32 bit single precision iteration,the 2D orthogonal polynomial model which is proposed in this paper has the best convergence,after the iteration,the fluctuation is small,and the results are almost coincident,and it is obviously better than the other two methods.Under the influence of different White Gaussian Noise,the predistortion performance of 2D orthogonal polynomial model is studied,it is also stable and has better performance of anti noise.Moreover,we can see through our studying and analysing,the coefficients of 2D orthogonal polynomial predistortion model than Legendre modified polynomial model reduced by more than half,and reducing the complexity of the predistortion system.In addition,it obtain a good linearity performance when verify the predistortion performance in the DPD test platform.