摘要
卫星通信中放大器使用的行波管(TWT)会引起信号的非线性畸变,对通信质量造成较大影响。为解决这种不利影响,在放大器前端采用非线性预失真,能较理想地消除放大器的非线性。由于神经网络能够对非线性函数进行较好地拟合,可以将其引入预失真器的设计。为简化神经网络中的LM算法,提出了一种采用LMS算法的系统模型,并建立新的自适应预失真器的结构模型,大大降低计算复杂度,有利于系统性能的提高。仿真表明,采用含有一个隐层(9个神经元)的神经网络模型来设计预失真器,能够达到较好的预校正效果。
The travel-wave-tube(TWT) used in the HPA in satellite communications may unexpectedly cause nonlinearity to the signals.A pre-distortion will reduce these distortions.The pre-distorter is based on a feed-forward neural network.To simplify the traditional Levenberg-Marquardt(LM) Algorithm for neural-network,a new least mean square(LMS) algorithm based model was proposed in which way the complexity of the algorithm was reduced.Simulation results confirmed that the proposed pre-distorter based on neural network consisting with one hidden layer and nine neurons produces a good performance improvement of quality of the transmission.
出处
《解放军理工大学学报(自然科学版)》
EI
北大核心
2011年第6期569-573,共5页
Journal of PLA University of Science and Technology(Natural Science Edition)
基金
江苏省自然科学基金资助项目(BK2009057)
关键词
神经网络
预失真
高增益放大器
行波管放大器
最小均方算法
neural network
pre-distortion
HPA(high-power amplifier)
TWTA(travel-wave-tube amplifier)
LMS