摘要
本文主要研究记忆多项式数字预失真技术。首先,对目前存在的多种线性化技术做综合的比较,结论显示预失真线性化技术具有明显的优越性。随后,本文构造一种基于直接学习结构(DLA)的记忆多项式预失真系统,其系数估计通过自适应算法实现,本文在传统的LMS算法的基础上推导出改进的LMS算法。文章最后利用64位QAM信号进行仿真验证,仿真结果表明:改进算法的稳定度有明显改善,精确度也有显著的提升,与本文构建的预失真系统相结合能够更好的消除失真。
Digital Pre-distortion of memory polynomial system is considered mainly in this paper. The pre-distortion is first constructed as a memory polynomial system and its coefficients are estimated based on the Direct Learning Architecture（DLA） approach. The Nonlinear Filter-x Least Mean Squares algorithm is derived assuming that the memory polynomial system is identified. The simulation results show that the memory polynomial pre-distorter can effectively reduce the distortion.
出处
《电子测试》
2011年第7期24-26,49,共4页
Electronic Test
