一种基于互相关序列的功放行为模型建立方法  

A Method of PA Behavioral Modeling Based on Cross-Correlation Sequence

作　　者：王华东[1] 何松柏[1] 鲍景富[1] 

机构地区：[1]电子科技大学电子工程学院,成都610054

出　　处：《电子与信息学报》2009年第5期1201-1204,共4页Journal of Electronics & Information Technology

摘　　要：该文提出了一种基于功率放大器行为模型误差信号与输入信号互相关的行为模型时延估计算法,并在此基础上讨论了如何动态建立放大器行为模型的方法。采用这种方法建立非均匀时延的记忆多项式(Memory Polynomial,MP)行为模型。仿真结果显示该文提出的时延估计算法能够有效地寻找到输出信号中的主要时延分量,从而将传统MP模型的时延长度由6减少到2,同时模型NMSE只降低不到3dB,表明这种行为模型建立方法能够很好地在模型复杂度与模型精度之间寻求平衡。In this paper, a time delay estimate algorithm based on analysis of cross-correlation sequence of model error and input signal is proposed, and a method to build behavioral model of power amplifiers dynamically depend on the new delay estimate algorithm is discussed. Using this method, a non-union delay Memory Polynomial（MP） model is built. Simulation results show the new algorithm can find main delays in output signal effectively and reduce the delay length of traditional MP model from 6 to 2 at the cost of 3 dB increase of model＇s NMSE, which means the algorithm can achieve a good balance between model complexity and its precision.

关 键 词：功率放大器 行为模型 记忆效应 互相关序列 

分 类 号：TN722[电子电信—电路与系统]