切比雪夫函数型连接神经网络在信道均衡中的应用

本文提出一种基于切比雪夫函数型连接神经网络(CFLNN)的信道均衡方法。传统的前馈神经网络虽然能有效地解决信道均衡的问题,但具有计算复杂度过高,收敛速度慢等缺点。函数型连接神经网络通过对输入模式进行非线性扩展,可以不必使用隐层而不降低整体性能,从而极大简化了网络结构。同时,神经网络的学习方法得以简化,提高了收敛速度。本文采用可变尺度共扼梯度下降法(SCG)对该函数型连接网络进行训练。仿真结果表明了用切比雪夫函数型连接神经网络解决信道均衡问题的有效性。     

一类输入元素非线性连接的反馈式神经网络及其联想能力分析

本文研究了输入元素非线性连的妆的反馈式神经网络，文中以二阶非线性连接为例给多拓扑结构，导出了能够实现模式平移不变识别的学习方法，并借助于等权类的概念把不变识别条件建造于网络权结构之中，同时降低了网络连接复杂度。     

复值Hopfield神经网络盲恢复8PSK信号

传统的神经网络恢复MPSK信号时,因接收信号是复值的,而采用将信号实部与虚部分开盲恢复,该类方法仍存在无法解决的多值信号盲恢复问题.本文利用接收数据正交补投影和发送信号属于有限字符集的先验知识基础上,在异步调整模式下,设计出复值空间的激励函数,将其引入神经网络的能量函数中,推出了一种基于离散复值Hopfield神经网络...     

基于输入扩展改进的BP网络及其在遥感图像分类中的应用

提出了一种基于输入模式扩展的神经网络改进方法,并和Levenberg-Marquardt优化的BP网络(LMBPN)进行了对比。通过二阶内积或切比雪夫多项式等非线性函数,把输入向量映射到更高维的模式空间,可以增强样本的可分性。Iris数据和遥感图像分类实验表明,输入模式扩展的神经网络改进方法可以进一步加快收敛速度,改进模式分类效果。     

一种MIMO系统中的快速广义复球形解码算法

该文提出了一种广义复球形解码算法。它能处理多输入多输出系统(MIMO)中发送天线多于接收天线的情形,并能同时检测具有格型结构和不具有格型结构的二维空间星座信号。该算法对信号矢量的超定部分进行优化搜索,从而避免了穷尽搜索的高复杂度。仿真结果表明该广义复球形解码算法的复杂度明显低于采用穷尽搜索策略的复杂度。     

基于权重双核模型的宽带功放行为建模方法

针对宽带功放行为建模和畸变补偿问题,提出了一种新的权重双盒模型.该模型采用无记忆子模型和记忆子模型权重级联的方法分别对宽带功放的静态非线性和动态非线性进行级联建模.首先,利用无记忆子模型对输入信号幅值进行压缩,降低静态畸变部分模型的拟合误差,并在子模型输出端进行解权值,保证功放的饱和驱动特性;接着,构建权重记忆子模型,利用信号幅值相关的权值函数动态的调整高低功率动态畸变子模型的权重比例.实验结果表明,在不同信号驱动情况下,本文方法在降低计算复杂度的同时,保持了更好的建模精度和畸变补偿能力.     

RBF神经网络在表面粗糙度光纤传感器中的应用

本文提出了以径向基函数（RBF）神经网络处理表面粗糙度光纤传感器输出信号的方法,将传感器的输出信号及作为光源的激光强信号同时加在RBF神经网络的输入端,利用RBF神经网络能够以任意精度逼近非线函数地能力的优点,同时实现对传感器的非线性补偿及减轻激光器输出光强变化带来的影响,采用这种方法的表面粗糙度光纤传感器,降低了对激光器输出功率稳定性的要求,具有测量范围大,精度高的特点。     

基于类脑脉冲神经网络的光纤油气管道监测

为了提高光纤油气管道监测效果,采用类脑脉冲神经网络算法。首先通过累加平均法对光纤传感器振动信号进行消噪;接着利用小波包对振动信号频带分解,将信号在频域上平均分为８个频段,把各频段的能量占比作为神经网络训练输入;然后基于脉冲响应对类脑神经元设计,神经元连接的强度与前、后神经元激活时间差函数关系非线性设计,振动输入数据与神经元脉冲转换;最后给出了算法流程。实验仿真显示本文算对单一振动信号识别准确率在9587左右,两种振动混合信号识别准确率在9052左右,指标优于其他算法,同时定位误差小于其他算法。     

一种前馈神经网络的快速学习算法

本文提出一种前馈神经网络的快速学习算法。与传统的BP方法相比,本算法有两个改进之处,一是同时将网络的非线性输出误差与线性输出误差作为待优化的目标函数,二是改进了学习过程中误差的反向传播因子。仿真结果表明,使用本文的算法训练前馈神经网络,计算复杂度略高于BP算法,但学习速度却有显著的提高。     

一种基于小波-神经网络的模拟电路故障诊断方法

小波分析具有数据压缩和特征提取的特性,神经网络具有非线性映射和学习推理的优点。结合两者的特点,提出了一种基于小波与神经网络的模拟电路故障诊断方法,该方法用小波变换对电路响应信号进行特征提取,从而简化神经网络的结构,降低计算的复杂度,加快了训练速度。对实例仿真表明,该法能有效地对模拟电路进行故障诊断。     

Functional link neural network cascaded with Chebyshev orthogonal polynomial for nonlinear channel equalization

Nonlinear intersymbol interference (ISI) leads to significant error rate in nonlinear communication and digital storage channel. In this paper, therefore, a novel computationally efficient functional link neural network cascaded with Chebyshev orthogonal polynomial is proposed to combat nonlinear ISI. The equalizer has a simple structure in which the nonlinearity is introduced by functional expansion of the input pattern by trigonometric polynomial and Chebyshev orthogonal polynomial. Due to the input pattern and nonlinear approximation enhancement, the proposed structure can approximate arbitrarily nonlinear decision boundaries. It has been utilized for nonlinear channel equalization. The performance of the proposed adaptive nonlinear equalizer is compared with functional link neural network (FLNN) equalizer, multilayer perceptron (MLP) network and radial basis function (RBF) along with conventional normalized least-mean-square algorithms (NLMS) for different linear and nonlinear channel models. The comparison of convergence rate, bit error rate (BER) and steady state error performance, and computational complexity involved for neural network equalizers is provided.     

Complex-valued neural networks with adaptive spline activationfunction for digital-radio-links nonlinear equalization

In this paper, a new complex-valued neural network based on adaptive activation functions is proposed. By varying the control points of a pair of Catmull-Rom cubic splines, which are used as an adaptable activation function, this new kind of neural network can be implemented as a very simple structure that is able to improve the generalization capabilities using few training samples. Due to its low architectural complexity (low overhead with respect to a simple FIR filter), this network can be used to cope with several nonlinear DSP problems at a high symbol rate. In particular, this work addresses the problem of nonlinear channel equalization. In fact, although several authors have already recognized the usefulness of a neural network as a channel equalizer, one problem has not yet been addressed: the high complexity and the very long data sequence needed to train the network. Several experimental results using a realistic channel model are reported that prove the effectiveness of the proposed network on equalizing a digital satellite radio link in the presence of noise, nonlinearities, and intersymbol interference (ISI)     

Blind Unitary Prewhitening with a Real-Valued Eigendecomposition

Prewhitening is a standard step for the processing of noisy signals. Typically, eigenvalue decomposition (EVD) of the sample data covariance matrix is used to calculate the whitening matrix. From a computational point of view, an important problem here is to reduce the complexity of the EVD of the complex-valued sample data covariance matrix. In this paper, we show that the computational complexity of the prewhitening step for complex-valued signals can be reduced approximately by a factor of four when the real-valued EVD is used instead of the complex-valued one. Such complexity reduction can be achieved for any axis-symmetric array. The performance of the proposed procedure is studied in application to a blind source separation (BSS) problem. For this application, the performance of the proposed prewhitening scheme is illustrated by means of simulations, and compared with the conventional prewhitening scheme. Among a number of BSS methods which use prewhitening, the second-order blind identification procedure has been adopted in this paper.     

一种基于复数残差网络的通信辐射源个体识别方法

在复杂电磁环境的通信辐射源个体识别任务中,针对传统特征提取识别方法分类效果不佳和低信噪比环境下基于实数神经网络的方法识别准确率不高的问题,本文提出了一种基于复数残差网络的通信辐射源个体识别方法。将实际采集的I路和Q路电台数据组合成复数作为输入,根据电台数据集特点选取复数初始化方法、复数激活函数,以改进的复数残差块为基础构建复数残差网络,进一步调整和优化网络结构并运用到辐射源个体识别任务中。通过实验证明,相比于实数残差网络和人工特征提取方法,复数残差网络的性能更优,并且在低信噪比的条件下,基于复数残差网络的方法鲁棒性更强。      

基于泛函连接网络和差分进化算法的后非线性混叠信号盲分离方法

把后非线性混叠信号盲分离的分离系统用泛函连接网络来建模,对分离系统的输出应用高阶统计量独立性准则作为测度,然后利用差分进化算法对泛函连接网络的权值进行学习,从而获得了一种后非线性混叠信号盲分离算法。由于泛函连接网络是一种单层神经网络,具有学习参数少、收敛速度快和非线性逼近能力强的特点;而差分进化算法控制参数少、易于选择、具有全局寻优能力和快速的收敛特性;因而与其它的后非线性混叠信号盲分离方法相比,该文提出的分离算法具有计算简单、收敛速度快、较高的精度和稳定性好的特点。仿真结果显示了这种方法是可行和有效的。     

Complex-Valued Recurrent Neural Network with IIR Neuron Model: Training and Applications

In this paper, based on the digital filter theory and approach, a new algorithm for training a complex-valued recurrent neural network, is proposed. Each recurrent neuron is modeled as an infinite impulse response (IIR) filter. The network weights are updated by optimizing the IIR filter coefficients, and the optimization is based on the layer-by-layer optimizing procedure as well as the recursive least-squares method. The performance of the proposed algorithm is demonstrated with application to a complex communication channel equalization. Our approach provides a new way to perform fast training of complex-valued recurrent neural networks.     

A factor graph approach to link loss monitoring in wireless sensor networks

The highly stochastic nature of wireless environments makes it desirable to monitor link loss rates in wireless sensor networks. In a wireless sensor network, link loss monitoring is particularly supported by the data aggregation communication paradigm of network traffic: the data collecting node can infer link loss rates on all links in the network by exploiting whether packets from various sensors are received, and there is no need to actively inject probing packets for inference purposes. In this paper, we present a low complexity algorithmic framework for link loss monitoring based on the recent modeling and computational methodology of factor graphs. The proposed algorithm iteratively updates the estimates of link losses upon receiving (or detecting the loss of) recently sent packets by the sensors. The algorithm exhibits good performance and scalability, and can be easily adapted to different statistical models of networking scenarios. In particular, due to its low complexity, the algorithm is particularly suitable as a long-term monitoring facility.     

基于复值深度神经网络的目标检测方法

针对雷达在检测概率要求严苛的多旁瓣干扰复杂场景下使用传统目标检测方法无法满足需求,性能有待进一步提升的问题,本文提出了一种基于多通道复值深度神经网络的雷达目标检测方法。传统脉冲体制阵列雷达的恒虚警率目标检测通常在和通道进行,在对回波信号进行空域相参预处理过程中获得了相参积累的同时丢失了阵元间的相位信息,而实际上目标回波在阵元间存在着一定的相位关系。本文利用神经网络强大的拟合能力和分类能力,将目标检测视为二元分类问题,设计复值深度神经网络深入挖掘目标与背景在不同阵元间的幅度及相位信息差异,从而在传统目标检测和通道-距离-多普勒空间的更前端更好地区分目标与背景的差异,提升了雷达目标检测性能。实验结果显示,所提方法在存在大量旁瓣干扰的场景下,相较传统方法具有更好的检测性能表现和抗干扰能力,且在杂波环境中也有良好的表现。      

一种快速复值自适应滤波算法

本文提出了一种基于欧几里德方向集方法的复值快速自适应滤波算法，它的计算复杂度为 Ｏ（Ｎ），并在理论上证明了该算法的稳定性。它的性能通过将其应用于自适应ＦＩＲ滤波中来加以说明，计算机仿真结果表明了该算法具有较快的收敛速度，这为自适应复值滤波提供了一种新方法。