基于分段线性函数的功放模型及数字预失真应用

发布时间：2018-04-22 06:19

本文选题：功率放大器 + 数字预失真　；参考：《计算机工程与科学》2017年11期

【摘要】：为了克服通信系统中功率放大器的非线性和记忆效应,数字预失真技术成为研究的热点。提出一种基于分段线性函数的多项式模型,与广义记忆多项式模型相比,我们把多项式中的高阶项转换为分段求和项,消除了高阶相乘带来的不稳定性,同时由于分段阈值的存在,该模型的适用性和稳定性均有所提高。把功放模型应用于数字预失真结构中的实验结果表明:与广义记忆多项式模型相比,分段线性函数模型所需系数要少40%,邻信道功率比提高约1dB,归一化均方误差提高约8dB,因此该模型在数字预失真方面具有较好的效果。
[Abstract]:In order to overcome the nonlinear and memory effects of power amplifiers in communication systems, digital predistortion technology has become a hot topic. A polynomial model based on piecewise linear function is proposed. Compared with the generalized memory polynomial model, we convert the higher order term in the polynomial to the piecewise sum term, which eliminates the instability caused by higher order multiplication. At the same time, the applicability and stability of the model are improved due to the existence of segmental threshold. The experimental results of applying the power amplifier model to the digital predistortion structure show that compared with the generalized memory polynomial model, The piecewise linear function model requires less than 40 coefficients, the power ratio of adjacent channel is increased by about 1 dB, and the normalized mean square error is increased by about 8 dB, so the model has a better effect on digital predistortion.
【作者单位】：天津大学电子信息工程学院;
【分类号】：TN722.75

【相似文献】