基于自适应预失真器的BLMS算法研究

Research on BLMS Algorithm Based on Adaptive Predistorter

由于功率放大器的特性随温度,电压等因素的变化而改变,为了保证预失真功放系统稳定工作,预失真系统的自适应性能就显得非常重要。本文首先基于记忆多项式构造预失真器,采用间接学习结构设计预失真系统,然后根据步长参数与误差,输入之间的非线性关系同时引入动量项提出了一个新的变步长(block least-mean-square,BLMS)算法。仿真表明,该算法用于功放预失真系统能够明显改善预失真收敛速度与精度。频谱密度改善30dBm。

The characteristic of the radio frequency power amplifier varies with many kinds of factors,such as the environment temperature,power supply,etc.In order to guarantee the steady operation of the predistortion power amplifier,the adaptive performance of the predistortion system seems very important.In this paper,the predistorter is constructed based on memory polynomial.The indirect learning architecture is adopted to design the predistortion system.According to the non-line relation ship between the step size and the error,at the same time momentum factor is introduced we propose a new function of variable step size.The results shows this algorithm has a more perfect effect in improving the speed and Accuracy of the predistortion amplifier.