Research On Digital Predistortion Of RF Power Amplifier Based On Volterra Series

With the rapid development of the mobile communication, in order to improve the utilization of the limited spectrum resources, non constant envelope modulation modes and transmission schemes are widely used in the modern communication systems. Because the signals of these modulation modes have the features of high envelop fluctuating and high peak-to-average ratio, during these signals pass through power amplifiers(PAs), a large amount of distortion will be produced. This put forward higher requirements of the linearity of PAs. Meanwhile, along with the bandwidth expansion of the communication systems, the memory effect of the PAs is also more and more obvious. In order to improve the communication quality and the efficiency of PAs, a linearization technique is required. Digital predistortion technology which based on Volterra series attracts more attentions due to its good linearity performance, high stability, low complexity and low cost.This paper mainly studies the behavioral modeling which based on Volterra series and predistortion methods of RF PAs. The main works and innovation points are as follows.Firstly, the principle and classification of the predistortion technology are elaborated, the existing behavior models of PAs are introduced, the predistortion learning structures are also introduced.Secondly, this paper proposed a modified and effective dynamic deviation reduction model. The modified DDR model was proposed based on the advantages of both the models of MP and DDR. The simulation and experimental results show that the modified model can achieves higher accuracy and better efficacy with less number of coefficients.Finally, the number of coefficients and complexity of the traditional Volterra series model and the simplified Volterra series model are analyzed deeply. We take the memoryless nonlinearity and the memory nonlinearity of PAs into consideration separately, based on these nonlinear components have different effects on the performance of models. This paper proposed a new and high robustness simplifiedVolterra series model(MSV). The experimental results show that the MSV model has advantages in complexity, accuracy and linearity.