基于RLS算法的记忆多项式预失真技术分析

自适应数字基带预失真技术是功放线性化技术中很有前景的技术之一。为了能够使数字基带预失真系统快速收敛且便于工程应用,介绍了一种基于迭代最小二乘(RLS)算法的记忆多项式预失真技术,并论述了其基本原理和系统结构,最后从邻信道功率比(ACPR)、星座图以及误码率3个方面进行了仿真,结果表明基于该算法的预失真技术不仅可以克服传统LMS预失真算法收敛速度慢的缺点,而且便于结合QR分解等方法来硬件实现,预失真性能更优。

Analysis of Memory Polynomial Predistortion Technique Based on RLS Algorithm

Adaptive digital baseband predistortion technique is one of the techniques with the most potential in the linearity of power amplifiers. In order to make the digital baseband predistortion system converge rapidly, this paper introduces a memory polynomial predistortion technique based on RLS algorithm and expounds its basic principles and structure, finally makes a simulation in aspects of adjacent channel power ratio, constellation and bit error. It is confirmed by simulation that this predistortion technique can not only overcome the disadvantage of slow convergence of LMS predistortion algorithm, but also be implemented conveniently by QR decomposition in the real hardware implementation as well as present better performance.