【摘要】 提出用神经网络的方法来实现功放的自适应预失真模型。它利用BP神经网络的函数逼近能力,来学习功放预失真器的AM/AM、AM/PM特性函数,以抵消由于功放非线性引起的信号失真和交扰;同时,也通过自适应地调整幅度和相位两个神经网络的权、阈值,来跟踪放大器的特性变化。仿真结果证实了基于神经网络的预失真模型的有效性和低复杂性。更多还原

【Abstract】 This paper presents a novel method of learning the AM/AM、AM/PM characteristics of the Amplifier抯 predistorter, using BP neural network抯 functional approximation property. It can cancel thesignal distortion and intermodulation caused by the amplifier抯 nonlinearity. At the same time, thismethod allows the tracking of changes in the amplifier抯 characteristics through adaptively adjusting the weights and thresholds of the two neural networks. Simulation results demonstrate the validity and low complexity of the predistortion model based on neural network.更多还原