摘要
预失真技术是克服功率放大器非线性失真的一种非常有效的方法.许多研究者从数学建模的思想角度出发,建立了关于功放与预失真的各类级数模型,以实现线性增益输出.但是在应用通信系统中,这些模型不能有效地动态实现功放效率尽可能高的要求.在和记忆多项式模型的基础上,创新地运用多目标规划的方法优化功放的预失真模型.在满足一定的线性化输出的同时,尽可能地得到更高的功放效率,而且能同时计算出增益系数与模型参数.对于数据量为1000的无记忆功放得到增益倍数为1.82649,预失真模型的NMSE值为-41.5633.数据量为73920的有记忆功放得到增益倍数为9.5969,预失真模型的NMSE值为-24.3682.
Predistortion technique is a highly effective method to overcome the nonlinear distortion of power amplifier. From the perspective of idea of mathematical modeling, many researchers work out a variety of series model of power amplifier and predistorter to achieve the linear gain output. However, in the application of communication systems, these models can＇t satisfy dynamically require for power amplifier efficiency as high as possible. Based on sum memory polynomial model in this paper, we innovate to apply multi-objective programming for solving predistorter modeling to optimize the predistorter model of power amplifier. On the condition of satisfying certain linear output, we try to improve the power amplifier efficiency as high as possible, besides, we can work out the gain coefficient and the model parameter at the same time. For memoryless PA of 1000 data, the gain coefficient is 1.82649 with the NMSE of preditorter is -41.5633. For memory PA of 72930, the results respectively are 9.5969 and -24.6482.
出处
《数学的实践与认识》
CSCD
北大核心
2014年第21期152-161,共10页
Mathematics in Practice and Theory
基金
国家自然科学基金(71363043)
关键词
多目标规划
预失真技术
最小二乘法
和记忆多项式
线性化
multi-objective programming
predistortion technology
least square method sum memory polynomial
linearization
