适用于DLA结构的新型功率放大器数字预失真算法

数字预失真技术通过在功率放大器的前端级联数字预失真器来补偿功放的非线性失真,而直接学习结构(Direct Learning Architecture,DLA)作为一种广泛使用的自适应数字预失真算法结构,因其预失真性能优良,前景最为看好。传统DLA预失真算法需要求得功率放大器高精度模型,其结构复杂且计算量大。针对这一问题,文章引入功放的瞬时复增益(Instantaneous Complex Gain,ICG),提出了两种适用于DLA的新型功放预失真算法:ICG-LMS和ICGRLS,在确保线性化效果的同时,省略功放模型构建过程,大大简化了算法结构。仿真分析和实测结果表明,与传统适用于DLA的自适应预失真算法相比,两种新型预失真算法的运行时间分别减少35.8%和32.1%,而功放输出信号的邻道功率比与原始功放输出相比分别降低了14.8 dB和23.5 dB。

Novel DLA-Based Digital Predistortion Algorithm for Power Amplifiers

The digital predistortion technique compensates for the nonlinear distortion of the power amplifier by cascading a predistortion model at the front end of the amplifier. Direct learning architecture(DLA) is one of the most widely used predistorters. However, available adaptive preditorters using direct learning architecture needs to obtain the high-precision model of the power amplifier, which is complicated in structure and requires a lot of computation. In this paper, we first propose the instantaneous complex gain(ICG) to simplify the pre-distortion structure and reduce the computational complexity. Through the simulation and experimental verification, the running time of the new pre-distortion algorithm is reduced by 35. 8% and 32. 1%, and the power ratio of the two algorithms in the adjacent channel at high and low frequency points is reduced by 14. 8 dB and 23. 5 dB respectively compared with that without pre-distortion.