Application of Recurrent Wavelet Neural Networks to the Digital Communications Channel Blind Equalization

ＡｐｐｌｉｃａｔｉｏｎｏｆＲｅｃｕｒｒｅｎｔＷａｖｅｌｅｔＮｅｕｒａｌＮｅｔｗｏｒｋｓｔｏｔｈｅＤｉｇｉｔａｌＣｏｍｍｕｎｉｃａｔｉｏｎｓＣｈａｎｎｅｌＢｌｉｎｄＥｑｕａｌｉｚａｔｉｏｎ＊＊ＴｈｉｓｗｏｒｋｗａｓｓｕｐｐｏｒｔｅｄｂｙｔｈｅＣｌｉｍｂ...     

A Neural Network Approach for Signal Detection in Digital Communications

In this paper, we propose a new approach for signal detection in wireless digital communications based on the neural network with transient chaos and time-varying gain (NNTCTG), and give a concrete model of the signal detector after appropriate transformations and mappings. It is well known that the problem of the maximum likelihood signal detection can be described as a complex optimization problem that has so many local optima that conventional Hopfield-type neural networks fail to solve. By refraining from the serious local optima problem of Hopfield-type neural networks, the NNTCTG makes use of the time-varying parameters of the recurrent neural network to control the evolving behavior of the network so that the network undergoes the transition from chaotic behavior to gradient convergence. It has richer and more flexible dynamics rather than conventional neural networks only with point attractors, so that it can be expected to have much ability to search for globally optimal or near-optimal solutions. After going through a transiently inverse-bifurcation process, the NNTCTG can approach the global optimum or the neighborhood of global optimum of our problem. Simulation experiments have been performed to show the effectiveness and validation of the proposed neural network based method for the signal detection in digital communications.     

保密通信中稳健差拍同步方法的描述与验证

结合自适应信道盲均衡和离散混沌系统的稳健差拍同步提出了一种存在信道畸变时的离散混沌系统的稳健差拍同步方法.该方法把混沌误差同步作为能量函数,利用进化算法训练递归神经网络完成自适应信道盲均衡,均衡器的输出作为混沌接收器的激励信号,从而在完成信道盲均衡的同时混沌系统达到同步状态.此外,提出了一种利用离散混沌系统稳健差拍同步实现存在信道畸变时的保密通信的方法.     

离散混沌系统稳健差拍同步方法及其在保密通信中的应用

本文结合自适应信道盲均衡和离散混沌系统差拍同步提出了一种存在信道畸变时的离散混沌系统的稳健关拍同步方法。该方法把混沌误差同步作为能量函数,利用进化算法训练递归神经网络完成自适应信道盲均衡,均衡器的输出作为混沌接收器的激励信号,从而在完成信道盲均衡的同时混沌系统达到同步状态。此外,本文提出了一种利用离散混沌系统稳健差拍同步实现存在信道畸变时的保密通信方法。     

基于深度学习的图像描述研究

卷积神经网络（Convolution Neural Networks,CNN）和循环神经网络（Recurrent NeuralNetworks,RNN）在图像分类、计算机视觉、自然语言处理、语音识别、机器翻译、语义分析等领域取得了迅速的发展,引起了研究者对计算机自动生成图像描述的广泛关注。目前图像描述存在的主要问题有输入文本数据稀疏、模型存在过拟合、模型损失函数震荡难以收敛等问题。文中使用NIC作为基线模型,针对数据稀疏问题,改变了基线模型中的文本one-hot表示,使用word2vec对文本进行映射,为了防止过拟合,在模型中加入了正则项和使用Dropout技术,并在词序记忆方面取得创新,引入联想记忆单元GRU,用于文本生成。在试验中使用AdamOptimizer优化器进行参数迭代更新。实验结果表明：改进后的模型参数减少且收敛速度大幅加快,损失函数曲线更加平滑,损失最大降至2.91,模型的准确率比NIC提高了接近15%。实验有效地验证了在模型当中使用word2vec对文本进行映射可明显缓解数据稀疏问题,加入正则项和使用Dropout技术可有效防止模型过拟合,引入联想记忆单元GRU能够大幅减少模型训练参数,加快算法收敛速度,进而提高整个模型的准确率。     

CPSN在信道均衡中的应用

信道均衡是用来消除码间串扰、对信道畸变进行补偿,从而在接收端正确的重建发送信号的滤波方法。由于信道均衡可以看作一个模式分类问题,恰恰神经网络具有良好的模式分类特征,因此针对复信道可采用CPSN进行二进制自适应信道均衡。仿真结果表明,该算法对于严重码间串扰及适度非线性畸变的复信道效果良好。     

应用粒子滤波器实现混沌通信系统的盲信道均衡

粒子滤波器(Particle filter,PF)是一种结合重要性权重抽样的序贯蒙特卡罗方法,能够应用到任意状态空间模型,并且能较好地估计经过非线性变化后的随机变量的统计特性.本文应用粒子滤波器和信号建模技术研究混沌通信系统的盲信道均衡问题,发展基于混沌的通信系统的盲均衡技术.仿真结果证实了,当Logistic映射作为混沌发生器和通信场景为固定参数与时变衰落信道时,该盲信道均衡器与基于扩展卡尔曼滤波算法的盲均衡器和基于无先导变换的自适应盲均衡器相比,有较好的均衡实现.此外,利用本文的盲均衡算法,实现了一种混沌调制通信系统的解调.     

Performance analysis of A dynamic channel allocation technique for satellite mobile cellular networks

This paper deals with an efficient dynamic channel allocation (DCA) technique suitable for applications in mobile satellite cellular networks. A cost function is defined to allow an optimum selection of channels to be allocated on demand. A mobility model suitable for low earth orbit (LEO) satellite systems is presented. The performance of the novel DCA technique in terms of call blocking probability has been derived by simulations. The obtained results are compared with those achieved by a fixed channel allocation (FCA) technique to show a better behaviour.     

RLS Algorithm with Variable Fogetting Factor for Decision Feedback Equalizer over Time-Variant Fading Channels

In a high-rate indoor wireless personal communication system, the delay spread due to multipath propagation results in intersymbol interference (ISI) which can significantly increase the transmission bit error rate (BER). Decision feedback equalizer (DFE) is an efficient approach to combating the ISI. Recursive least squares (RLS) algorithm with a constant forgetting factor is often used to update the tap-coefficient vector of the DFE for ISI-free transmission. However, using a constant forgetting factor may not yield the optimal performance in a nonstationary environment. In this paper, an adaptive algorithm is developed to obtain a time-varying forgetting factor. The forgetting factor is used with the RLS algorithm in a DFE for calculating the tap-coefficient vector in order to minimize the squared equalization error due to input noise and due to channel dynamics. The algorithm is derived based on the argument that, for optimal filtering, the equalization errors should be uncorrelated. The adaptive forgetting factor can be obtained based on on-line equalization error measurements. Computer simulation results demonstrate that better transmission performance can be achieved by using the RLS algorithm with the adaptive forgetting factor than that with a constant forgetting factor previously proposed for optimal steady-state performance or a variable forgetting factor for a near deterministic system.     

实现标量混沌信号同步控制的神经网络方法

本文给出了一种利用线性输出神经网络实现标量混沌信号同步控制的方法。该方法利用线性输出神经网络构造被控混沌系统的模型，并基于Lyapunov理论与非线性系统控制方法，设计出神经网络权值变化规律与非线性反馈控制器，使神经网络模型的标量输出能大范围同步于给定的标量混沌信号。理论分析与计算机模拟结果都证实了这种方法的有效性。     

基于误差反向传播算法的OFDM系统频域均衡

提出用一种基于神经网络的均衡方法,并用误差反向传播(BP)算法对多层前馈网络进行训练.分析和数值仿真结果表明,基于BP网络的频域均衡方法有较好的均衡效果,并且复杂度不高.     

一种基于神经网络的混沌序列产生方法

应用具有全局最优的ＢＰ改进算法,建立了产生混沌序列的神经网络模型（ＣＧＮＮ）。该模型产生序列随机性良好,序列更换调整容易。对ＣＧＮＮ系统的抗破译性能进行了分析,结果表明,该模型产生的混沌序列确可以作为最优加密密匙及扩频码。     

一种新的混沌扩频序列产生方法

本文提出了一种新的混沌扩频序列产生方法。该方法基于神经网络的强大学习能力和副近非线性函数能力,应用具有全局最优的ＢＰ改进算法通过训练学习建立起具有混沌性态的优化神经网络模型,利用网络权值调整的灵活性来产生混沌扩频序列。计算机仿真结果表明,该模型产生的混沌扩频序列调整更容易,比基于单一混沌映射能产生更多符合扩频通信要求的扩频序列。     

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。     

混沌同步和混沌通信研究的新进展与新尝试

对目前现有的几种混沌同步方法和混沌通信方案及其最新发展进行较全面的评述,着重于各种方案的比较、对要点的论述,并且介绍了一些以往未引起普遍注意的重要成果。     

利用EM算法估计隐含观测量的回馈神经网络学习新方法

研究回馈神经网络(RNN)参数估计的新方法.利用隐含观测量,将复杂RNN的训练分解为线性输出层和多个单隐元的参数估计.基于每个隐元激励函数的多点线性近似,RNN可利用统计混合专家网络模型(ME)描述,从而将RNN的参数估计转化为包含隐含观测量的线性系统的最大似然估计问题,最后利用期望最大化(EM)算法获得RNN的隐含观测量及其参数估计.     

Design and performance evaluation of a full-duplex operating receiver for time-hopping UWB

One of the most attractive features of time-hopping ultra-wide-band (UWB) transmission, largely ignored in the literature so far, is the possibility to operate in full-duplex mode, thanks to its very low duty-cycle. This allows a terminal to transmit and receive within the same time frame and frequency band, yielding a considerable saving of time and band resources at the radio-resource-management layer. In this paper, we propose a methodology to design channel estimation/synchronization and demodulation/decoding algorithms for a low-complexity receiver operating in full-duplex mode. The leading idea is simply to avoid, at reception, time intervals that correspond to pulse transmission. We also evaluate the performances of such a system through simulations over realistic propagation channels. Tomaso Erseghe was born in Valdagno, Italy, in 1972. He received the laurea degree and the PhD in Telecommunication Engineering from the University of Padova, Italy, respectively in 1996 and in 2002. From 1997 to 1999 he worked as an R&D Engineer at Snell & Wilcox, a British broadcast equipment manufacturer, in the areas of image restoration and motion compensation. From 2003 he is an Assistant Professor (Ricercatore) at the University of Padova. His research interests include fractional Fourier transforms, lossless encoding algorithms, and ultra-wideband transmission systems. Nicola Laurenti was born in 1970, in Adria, Italy. He graduated from the University of Padova with a laurea in Electrical Engineering in 1995, with a thesis on Image Reconstructions from Projections, and obtained a Ph.D. in Electrical and Telecommunications Engineering from the same University in February 1999, with a thesis on Implementation Issues in OFDM Systems. Since 2001 he is an Assistant Professor at the Department of Information Engineering, University of Padova. His research interests mainly focus on digital communications, especially multicarrier modulation and ultra wide band transmission, but also include signal theory, and the processing of audio and biomedical signals.