基于神经网络的接收机宽带非线性行为建模

doi:10.16180/j.cnki.issn1007-7820.2022.08.001

摘要/Abstract

摘要：

为了预测复杂电磁环境下接收机的非线性效应,文中基于实数时延径向基函数神经网络,构建了具有记忆效应的接收机非线性神经网络模型。分别采用K-均值聚类算法和正交最小二乘法对模型的隐含层中心和权值进行选取和学习,并用接收机的输入输出实测数据对模型进行训练。通过宽带信号的同相和正交两个分量对模型进行验证。模型仿真结果与实测数据相吻合,模型的归一化均方误差可达-41.88 dB。该结果表明,所构建的神经网络模型具有较快的收敛速度、良好的建模精度和泛化能力。

关键词: 接收机, 非线性, 行为建模, 径向基函数, 神经网络, 记忆效应, 宽带信号, 泛化能力

Abstract:

In order to predict the nonlinear effect of receiver in complex electromagnetic environment, a nonlinear behavior model of receiver with memory effect is constructed based on real-value time-delay radial basis function neural network. The K-means clustering algorithm and the orthogonal least square method are respectively used to select and learn the center of the hidden layer and weight of the model, and the model is trained with the input and output measured data of the receiver. The model is verified by the in-phase and quadrature components of wideband signals. The simulation results are in good agreement with the measured data, and the normalized mean square errors of the model reaches -41.88 dB. The verification results show that the neural network model has fast convergence speed, good modeling accuracy and generalization ability.

Key words: receiver, nonlinear, behavioral modeling, radial basis function, neural network, memory effect, wideband signals, generalization ability

中图分类号:

TN85

刘国华,路宏敏,陈冲冲,李万玉,万健鹏. 基于神经网络的接收机宽带非线性行为建模[J]. 电子科技, 2022, 35(8): 1-6.

LIU Guohua,LU Hongmin,CHEN Chongchong,LI Wanyu,WAN Jianpeng. Wideband Nonlinear Behavior Modeling of Receiver with Neural Network[J]. Electronic Science and Technology, 2022, 35(8): 1-6.

图/表 9

图1

图2

图3

图4

图5

表1

图6

图7

图8

参考文献 19

[1]	胡向峰, 张艳花. 接收机的非线性分析[J]. 太赫兹科学与电子信息学报, 2008, 6(6):440-443.
	Hu Xiangfeng, Zhang Yanhua. Nonlinear analysis of receiver[J]. Journal of Terahertz Science and Electronic Information Technology, 2008, 6(6):440-443.
[2]	Carvalho N B. Intermodulation distortion in microwave and wireless circuits[J]. Liver International Official Journal of the International Association for the Study of the Liver, 2003, 35(2):660-672.
[3]	李培红. 无源器件的三阶互调研究及测试[J]. 电子科技, 2015, 28(9):112-114.
	Li Peihong. Third-order intermodulation of passive components research and testing[J]. Electronic Science and Technology, 2015, 28(9):112.
[4]	何军. 宽带射频功率放大器记忆效应的研究[D]. 成都: 电子科技大学, 2009.
	He Jun. Study on memory effect of wideband RF power amplifier[D]. Chengdu: University of Electronic Science and Technology of China, 2009.
[5]	Xu G, Liu T, Ye Y, et al. Generalized two-box cascaded nonlinear behavioral model for radio frequency power amplifiers with strong memory effects[J]. IEEE Transactions on Microwave Theory & Techniques, 2014, 62(12):2888-2899. doi: 10.1109/TMTT.2014.2365459
[6]	刘太君, 陈豪, 苏日娜, 等. 基于神经网络的宽带功放动态非线性行为建模[J]. 微波学报, 2020, 36(1):131-136.
	Liu Taijun, Chen Hao, Su Rina, et al. Neural networks based nonlinear dynamic behavior modelling for broadband power amplifiers[J]. Journal of Microwaves, 2020, 36(1):131-136.
[7]	Sappal A S. Simplified memory polynomial modelling of power amplifier[C]. Vancouver: International Conference and Workshop on Computing and Communication, 2015.
[8]	Bai E W, Tempo R. Representation and identification of non-parametric nonlinear systems of short term memory and low degree of interaction[J]. Automatica, 2010, 46(10):1595-1603. doi: 10.1016/j.automatica.2010.06.031
[9]	李敏玥. 基于多项式的接收机非线性行为建模方法研究[D]. 西安: 西安电子科技大学, 2020.
	Li Minyue. Research on receiver nonlinear behavioral modeling method based on polynomial[D]. Xi'an: Xidian University, 2020.
[10]	赵一鹤, 邵杰, 程永亮. 基于编码—解码模型的D类功率放大器行为建模[J]. 电子科技, 2020, 33(2):20-24.
	Zhao Yihe, Shao Jie, Cheng Yongliang. Behavior modeling of class-D power amplifier based on encoder-decoder model[J]. Electronic Science and Technology, 2020, 33(2):20-24.
[11]	Liu T, Hui M, Zhang Y, et al. RF power amplifier modeling and linearization with augmented RBF neural networks[C]. Nanjing:IEEE International Workshop on Electromagnetics:Applications and Student Innovation Competition, 2016.
[12]	Zhai J, Zhou J, Zhang L, et al. The dynamic behavioral model of RF power amplifiers with the modified ANFIS[J]. IEEE Transactions on Microwave Theory and Techniques, 2009, 57(1):27-35. doi: 10.1109/TMTT.2008.2009085
[13]	Yan S, Zhang C, Zhang Q J. Recurrent neural network technique for behavioral modeling of power amplifier with memory effects[J]. International Journal of RF and Microwave Computer-Aided Engineering, 2015, 25(4):289-298. doi: 10.1002/mmce.20861
[14]	Rawat M, Rawat K, Ghannouchi F M. Adaptive digital predistortion of wireless power amplifiers/transmitters using dynamic real-valued focused time-delay line neural networks[J]. IEEE Transactions on Microwave Theory and Techniques, 2010, 58(1):95-104. doi: 10.1109/TMTT.2009.2036334
[15]	Zhao Z, Na W, Zhang Q. Multi-band behavioral modeling of power amplifier using carrier frequency-dependent time delay neural network model[C]. Seville:IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization for RF,Microwave and Terahertz Applications, 2017.
[16]	Tan H, Huang M. Numerical simulation of the nonlinear response of receiver mixer[C]. Suzhou:International Applied Computational Electromagnetics Society Symposium, 2017.
[17]	Shi N, Guan Y, Liu X. Research on k-means clustering algorithm: An improved k-means clustering algorithm[C]. Ji'an:The Third International Symposium on Intelligent Information Technology and Security Informatics, 2010.
[18]	郭海如, 李志敏, 万兴, 等. 一种基于随机GA的提高BP网络泛化能力的方法[J]. 计算机技术与发展, 2014, 24(1):105-108.
	Guo Hairu, Li Zhimin, Wan Xing, et al. A method of improving generalization for BP network based on random GA[J]. Computer Technology and Development, 2014, 24(1):105-108.
[19]	武鹏, 郭晓芸, 王海龙, 等. 基于卷积神经网络模型的情绪识别技术在语音质检中的应用[J]. 电子设计工程, 2021, 29(5):164-168.
	Wu Peng, Guo Xiaoyun, Wang Hailong, et al. Emotion recognition technique based on convolutional neural network model application in voice quality inspection[J]. Electronic Design Engineering, 2021, 29(5):164-168.

记忆深度		2	3	4	5
NMSE/dB	RVTDRBFNN	-35.96	-41.88	-40.45	-38.24
NMSE/dB	RVTDBPNN	-35.22	-34.57	-34.12	-35.36