基于非迭代算法和非直接学习结构的查询表TWTA预失真器

Look-up table predistorter based on non-iterative algorithm and indirect learning architecture for TWTA

提出了一种基于查询表和非直接学习结构的适用于行波管放大器(TWTA)的基带预失真器。通过引入线性增长的周期性训练序列和失真检测算法,实现了预失真器非迭代方法的参数捕获。与普通查询表预失真器相比,该预失真器不但能离线处理失真数据,而且不需要使用基于最小代价函数的迭代算法更新预失真器查询表中的内容,从而能在降低对系统处理速度要求的同时,回避了算法收敛性问题。仿真结果表明,所提预失真器可大大降低由TWTA非线性所引入的带内失真和频谱再生。

Based on look-up table （LUT） and indirect learning architecture, a baseband predistorter was proposed for travel wave tube amplifier （TWTA）. Non-iterative method for capturing parameters of the predistorter was realized by the means of linear increasing periodic training sequence and distortion estimation algorithm. Compared with the conventional predistorter based on LUT, the proposed predistorter could process distortion data offline without minimum cost function to update the LUT, which lowers the processing requirement of the predistortion system and avoids convergence problem of adaptive algorithm. Simulation results show that the predistorter can reduce in-band distortion and spectrum regrowth greatly introduced by the nonlinear effects of TWTA.