基于编码—解码模型的D类功率放大器行为建模

D类功率放大器具有优异的传输效率,属于开关类功放,其输出信号存在较大的非线性失真。对D类功率放大器进行行为建模时要同时考虑其非线性和记忆特性。文中将小波变换引入到编码—解码神经网络模型中,提出了小波编码—解码神经网络模型。使用基于门限循环单元的编码—解码模型和小波编码—解码模型进行D类功率放大器的行为建模。实验结果表明,文中提出的D类功率放大器行为模型相比于传统的Voterra-Laguerre模型而言,在信号的时域和频域都具有更高的精度。     

7～18GHz单片宽带大功率放大器

宽带 Ga AS MMIC功率放大器是雷达及微波宽带测试中的关键部件 ,在有关电子系统和微波通信中得到广泛应用。南京电子器件研究所最近研制成功的 7～ 1 8GHz宽带大功率放大器 ,具有比较优异的性能。该宽带单片功率放大器采用 76mm、0 .5 μmPHEMT MMIC工艺线完成。放大器的拓扑结构采用两级有耗匹配方式 ,双路信号在片内用 Lang桥直接合成 ,体积较小 ,从而实现高增益、大功率。宽频带大信号模型的建立是其中的关键 ,利用微波在片测试系统结合模型提取软件进行非线性模型的提取 ,并利用负载牵引系统对提取的模型进行验证、优化。利用…     

模糊格构造型形态神经网络

针对构造型形态神经网络(CMNN)决策函数的局限性,提出了一种模糊格构造型形态神经网络(FLCMNN);该模型在利用训练好的CMNN进行分类时,引入模糊格包容性测度计算测试样本属于各超盒的隶属度值.采用仿真数据集对提出的FL-CMNN模型进行了评价,并与原始的CMNN和传统的人工神经网络、支持向量机、最近邻分类器进行了对比;试验结果表明,FL-CMNN在测试精度上明显优于原始的CMNN,训练时间远远低于传统的神经网络和支持向量机,而分类精度丝毫不亚于传统的神经网络和支持向量机.     

基于模糊神经网络智能预测模型的设计与实现

针对智能决策支持系统中经常遇到的预测类问题,根据人工神经网络和模糊逻辑系统的各自特点,设计一种模糊神经网络模型,将模糊系统用类似于神经网络的结构表示,再用相应的学习算法训练模糊系统实现模糊推理.并对此模型进行预测验证和编程实现.     

基于神经网络的接收机宽带非线性行为建模

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。     

一种基于延迟线补偿记忆效应的模拟预失真器

模拟预失真器具有带宽宽、结构简单、功耗低和延时少等优点,满足第五代移动通信系统(5G)及超 5G 的功放线性化对大带宽、低功耗和低延时的要求。然而随着移动通信系统的发展,信号的带宽和调制度越来越 高,功率放大器的记忆效应影响也越来越强,而传统的模拟预失真器无法补偿功放的记忆效应。为了解决模拟预失 真电路的记忆效应补偿问题,文中提出了一种基于延迟线补偿记忆效应的肖特基二极管模拟预失真器(SDD-APD)。 该模拟预失真器采用不等长微带线作为延迟线,用来补偿功放的记忆效应。采用100 MHz 带宽5G 新无线电(NR) 信号对工作在3. 5 GHz 的AB 类功放进行测试,结果表明该模拟预失真器可以补偿功放的记忆效应,并能将功放的 非线性改善10 dB 以上。     

一种基于神经网络的图像复原方法

提出了一种基于BP神经网络的图像复原算法.在分析图像模糊机制的基础上,为了降低输入维数,该方法采用滑动窗口操作来提取特征,同时为了加快训练速度和改善网络复原效果,首先对图像进行边缘提取,对图像内边缘区域和平坦区域分别采用滑动窗口获得训练集.利用BP神经网络的学习能力,通过训练,建立含有退化信息(高斯模糊)的模糊图像和清晰图像之间的映射关系模型,利用该模型对模糊图像进行复原,得到的复原图像在视觉上和定量分析上都获得了较好的效果.     

改进BP神经网络的功放有记忆行为模型

提出了一种基于改进误差反向传播神经网络(IBPNN)的具有记忆效应功率放大器(PA)的行为模型。该模型在传统误差反向传播神经网络(BPNN)的基础上利用Levenberg-Marquardt(LM)学习算法和加入动量因子的训练算法更新BPNN的权值和阈值,与传统的BPNN相比只需要更少的训练次数就达到了更高的精度。20MHz带宽三载波WCDMA信号的时域和频域仿真都表明其具有良好的性能,并且由得到的功率放大器(PA)动态特性AM/AM和AM/PM可知,该模型可以很好地描述PA的记忆效应。最后,用16QAM调制的OFDM 20MHz带宽信号的实验证明了该模型具有普遍的适用性。     

记忆多项式数字预失真线性化逆E类功放

采用记忆多项式模型的数字预失真器,用于线性化逆E类射频功率放大器,从而获得具有高线性和高效率的射频放大系统,使得开关型的逆E类功率放大器可以适用于具有非恒包络的调制信号的发射。文中设计了一个工作于S频段的具有10W饱和功率的逆E类功率放大器,以具有5MHz信号带宽的单载波WCDMA信号作为测试信号,使用记忆多项式的预失真器对其进行线性化。实验表明,该记忆多项式预失真器能够很好地抑制逆E类功放的动态非线性引起的带外寄生频谱,可以使逆E类功放同时工作于高线性和高效率状态。     

基于神经网络的功率放大器预失真建模

提出一种神经网络结合分离信号对功率放大器预失真建模的方法。将输入/输出信号的线性与非线性部分分开处理,利用神经网络良好的逼近能力,采用LM算法,拟合出功率放大器特性曲线,进而建立预失真模型,使非线性功率放大器的输入/输出曲线整体呈线性化。在保证输出幅度限制和输出功率最大化的前提下,与未作信号分离的神经网络建模方法、多项式建模方法以及Saleh函数模型方法相比较,发现信号分离神经网络建模方法能得到较小的归一化均方误差和误差矢量幅度。仿真结果表明,采用信号分离神经网络对功率放大器及其预失真建模,整体线性化误差较小、精度高、效果更佳。     

The Dynamic Behavioral Model of RF Power Amplifiers With the Modified ANFIS

The adaptive neuro-fuzzy inference system (ANFIS) has been widely used for modeling different kinds of nonlinear systems including RF power amplifiers (PAs). The modified ANFIS (MANFIS) architecture is simpler than that of ANFIS, but with nearly the same performance for modeling nonlinear systems. In this paper, the MANFIS is applied to model RF PAs with memory effects. The simulation and experimental results both in the time and frequency domains show that this model has good modeling accuracy and the characteristics of faster convergence and lower computational complexity compared with the ANFIS model. The normalized mean squared errors of the MANFIS model are slightly lower than those of some other neural network models such as the real-valued time delay neural network, radial basis-function neural network, etc. Finally, the MANFIS model is successfully used in a digital predistortion system, which can provide over 10- dB adjacent channel leakage ratio improvement for three-carrier wideband code division multiple access signals.     

基于神经网络的宽带功放动态非线性行为建模

文章主要讨论了如何利用神经网络对宽带功放进行动态非线性行为建模的问题。首先简述了功放的动态非线性特性及行为建模的方法。然后回顾了基于实数时延前馈神经网络、径向基函数神经网络等浅层神经网络构建的功放动态非线性行为模型。在此基础上,针对5G/6G宽带功放具有更强的记忆效应的问题,重点分析了如何使用长短期记忆(LSTM)神经网络对功放的动态非线性进行精确的行为建模。最后展望了构建具有普适性的功放非线性行为模型将是5G/6G通信时代功放非线性建模的一个重要发展方向。     

Neuro-fuzzy modeling and control

Fundamental and advanced developments in neuro-fuzzy synergisms for modeling and control are reviewed. The essential part of neuro-fuzzy synergisms comes from a common framework called adaptive networks, which unifies both neural networks and fuzzy models. The fuzzy models under the framework of adaptive networks is called adaptive-network-based fuzzy inference system (ANFIS), which possess certain advantages over neural networks. We introduce the design methods for ANFIS in both modeling and control applications. Current problems and future directions for neuro-fuzzy approaches are also addressed     

Cheap Joint Probabilistic Data Association with Adaptive Neuro-Fuzzy Inference System State Filter for Tracking Multiple Targets in Cluttered Environment

The cheap joint probabilistic data association (CJPDA) with the adaptive neuro-fuzzy inference system state filter (ANFISSF) is presented for tracking multiple targets in the presence of low and high cluttered environments. The state update step of the CJPDA filter (CJPDAF) is realized with the ANFISSF instead of Kalman filter. The adaptive neuro-fuzzy inference system (ANFIS) has the advantages of expert knowledge of fuzzy inference system and learning capability of neural networks. A hybrid learning algorithm, which combines the least square method and the backpropagation algorithm, is used to identify the parameters of ANFIS. The tracks estimated by using the method proposed in this paper for different tracking scenarios are in very good agreement with the original tracks.     

ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM FOR COMPUTING THE PHYSICAL DIMENSIONS OF ELECTRICALLY THIN AND THICK RECTANGULAR MICROSTRIP ANTENNAS

A new method for calculating the patch length and width of rectangular microstrip antennas (MSAs) with thin and thick substrates, based on adaptive neuro-fuzzy inference system (ANFIS), is presented. The ANFIS is a class of adaptive networks which are functionally equivalent to fuzzy inference systems (FISs). It combines the powerful features of FISs with those of artificial neural networks (ANNs) to achieve a desired performance. The results of ANFIS are in excellent agreement with the experimental results available in the literature.     

Adaptive neuro-fuzzy inference system for computing the resonant frequency of electrically thin and thick rectangular microstrip antennas

A new method based on adaptive neuro-fuzzy inference system (ANFIS) for calculating the resonant frequency of rectangular microstrip antennas (MSAs) with thin and thick substrates is presented. The ANFIS has the advantages of expert knowledge of fuzzy inference systems (FISs) and learning capability of artificial neural networks (ANNs). A hybrid learning algorithm, which combines the least square method and the backpropagation algorithm, is used to identify the parameters of ANFIS. The resonant frequency results obtained by using ANFIS are in excellent agreement with the experimental results reported elsewhere.     

FPGA-implementation of dynamic time delay neural network for power amplifier behavioral modeling

In this paper, we propose new architectures for FPGA-implementation of a dynamic neural network power amplifier behavioral modeling. The real-valued time-delay neural network (RVTDNN) and the backpropagation (BP) learning algorithm were implemented on FPGA using Xilinx system generator for DSP and the Virtex-6 FPGA ML605 evaluation kit. Different RVTDNN architectures are proposed for various values of the number of hidden neurons, the activation function resolution, and the fixed-point data format. These architectures are evaluated and compared in terms of modeling performances and resource utilization using 16-QAM modulated test signal.     

用记忆型 BP 神经网络实现 HPA 预失真的算法研究简

在分析宽频带CMMB直放站高功率功放(HPA）特性的基础上,提出了一种可分离处理功放记忆效应和非线性的延时神经网络（FIR-NLNNN）模型。该模型以实数延时神经网络（RVTDNN）为基础,用Levenberg-Marquardt（LM）优化算法确定神经网络系数,在模型中新增参数w0,给出了LM算法的修改公式。接着在预失真神经网络系统中引入Bayesian机理消除LM算法的过拟合现象,构建CMMB数字直放站的间接学习预失真器,拟合HPA的非线性和记忆效应。结果表明：RVTDNN和FIR-NLNNN 2种预失真器均能显著提高系统性能,降低邻信道功率比30 dB左右。在保持均方误差（MSE）小于10?6的情况下,FIR-NLNNN结构的网络参数比RVTDNN结构减少了近50%,迭代过程中的乘法和加法次数约降低75%。     

Macrocell path loss prediction using artificial intelligence techniques

The prediction of propagation loss is a practical non-linear function approximation problem which linear regression or auto-regression models are limited in their ability to handle. However, some computational Intelligence techniques such as artificial neural networks (ANNs) and adaptive neuro-fuzzy inference systems (ANFISs) have been shown to have great ability to handle non-linear function approximation and prediction problems. In this study, the multiple layer perceptron neural network (MLP-NN), radial basis function neural network (RBF-NN) and an ANFIS network were trained using actual signal strength measurement taken at certain suburban areas of Bauchi metropolis, Nigeria. The trained networks were then used to predict propagation losses at the stated areas under differing conditions. The predictions were compared with the prediction accuracy of the popular Hata model. It was observed that ANFIS model gave a better fit in all cases having higher R² values in each case and on average is more robust than MLP and RBF models as it generalises better to a different data.     

Input Resistance Calculation for Circular Microstrip Antennas Using Adaptive Neuro-Fuzzy Inference System

This paper presents a new method based on adaptive neuro-fuzzy inference system (ANFIS) to calculate the input resistance of circular microstrip patch antennas. The ANFIS is a fuzzy inference system (FIS) implemented in the framework of an adaptive fuzzy neural network. It combines the explicit knowledge representation of FIS with learning power of neural networks. A hybrid learning algorithm based on the least square approach and the backpropagation algorithm is used to optimize the parameters of ANFIS. The input resistance results predicted by ANFIS are in excellent agreement with the experimental results reported elsewhere.