Application Research Of Digital Predistortion

Nowadays, wireless communication systems are evolving to support more users and provide higher transmission rate services within the limited frequency band. In order to ensure the efficient use of the spectrum, people proposed a variety of non-constant envelope modulation techniques, such as wideband code division multiple access, orthogonal frequency division multiplexing and so on. However, these modulation signals have the characteristics of high peak to average ratio, which can stimulate the inherent nonlinear distortion of power amplifiers (PAs), and affect the performance of the communication systems. To this end, it is necessary to linearize power amplifiers. At present, digital predistortion (DPD) has become the mainstream of RF power amplifier’s linearization technology. In this thesis, we focus on the research of wideband digital predistortion and compensation of nonlinear distortion in the transmitter.In traditional digital predistortion, the loop delay estimation of DPD system is a very important step. If the loop delay is not accurately estimated, the memory effect of the power amplifier will be overestimated. In this thesis, we first analyze the method of amplitude difference cross-correlation function and digital differentiator to estimate the loop’s integer delay, then we propose a new fractional delay estimation algorithm based on the 4 tap piecewise parabolic interpolation filter.In order to meet the demand of high data rate transmission, the wireless communication systems are developing in the direction of wideband, which brings great challenge to the analog to digital converters of the feedback path in digital predistortion systems. This thesis first analyzes the spectrum extrapolation method to realize DPD with under-sampling in the feedback path. After that, this thesis presents a new method of under-sampling DPD based on random demodulation technology. Simulation results show that the new method is well suitable for the application of wideband digital predistortion.The non-ideal characteristics of the hardware circuits of the quadrature modulator can influence the performance of the traditional DPD, which is more obvious in concurrent multi-band transmitters. Considering that the two-dimensional model is very complex to jointly compensate the non-ideal characteristics of the modulator and the concurrent multi-band PA distortion, this thesis proposes to use the orthogonal matching pursuit algorithm of compressed sensing to simplify the model. The simplified model can maintain the performance of the original model with relatively lower complexity.In addition, this thesis also proposes a new model for jointly compensating the non-ideal characteristics of the modulator and the concurrent multi-band PA distortion. Compared with the original models, experiment results show that the new model has the advantages of lower complexity and better linearization performance.