Congestion warning method based on the Internet of vehicles and community discovery of complex networks

The traffic congestion occurs frequently in urban areas, while most existing solutions only take effects after congesting. In this paper, a congestion warning method is proposed based on the Internet of vehicles(IOV) and community discovery of complex networks. The communities in complex network model of traffic flow reflect the local aggregation of vehicles in the traffic system, and it is used to predict the upcoming congestion. The real-time information of vehicles on the roads is obtained from the IOV, which includes the locations, speeds and orientations of vehicles. Then the vehicles are mapped into nodes of network, the links between nodes are determined by the correlations between vehicles in terms of location and speed. The complex network model of traffic flow is hereby established. The communities in this complex network are discovered by fast Newman(FN) algorithm, and the congestion warnings are generated according to the communities selected by scale and density. This method can detect the tendency of traffic aggregation and provide warnings before congestion occurs. The simulations show that the method proposed in this paper is effective and practicable, and makes it possible to take action before traffic congestion.     

多方数据共享环境下匿名化数据的安全性量化方法(英文)

This paper aims to find a practical way of quantitatively representing the privacy of network data. A method of quantifying the privacy of network data anonymization based on similarity distance and entropy in the scenario involving multiparty network data sharing with Trusted Third Party (TTP) is proposed. Simulations are then conducted using network data from different sources, and show that the measurement indicators defined in this paper can adequately quantify the privacy of the network. In particular, it can indicate the effect of the auxiliary information of the adversary on privacy.     

An effective network congestion control method for multilayer network

The congestion control problem in a single node network has been solved by the nonlinear feedback control method,which has been proven to be effective and robust for different router’s queue size.However,these control models are based on the single layer network architecture,and the senders and receivers are directly connected by one pair of routers.With the network architecture being more and more complex,it is a serious problem how to cooperate many routers working in the multilayer network simultaneously.In this paper,an effective Active Queue Management（AQM）scheme to guarantee the stability by the nonlinear control of imposing some restrictions on AQM parameter in multilayer network is proposed.The nonlinear control can rely on some heuristics and network traffic controllers that appear to be highly correlated with the multilayer network status.The proposed method is based on the improved classical Random Early Detection（RED）differential equation and a theorem for network congestion control.The theorem proposed in the paper proved that the stability of the fluid model can effectively ensure the convergence of the average rate to its equilibrium point through many routers in multilayer network.Moreover,when the network capacity is larger,the proposed scheme can still approach to the fullest extensibility of utilization and ensure the stability of the fluid model.The paper reveals the reasons of congestion control in multilayer network,provides a theorem for avoiding network congestion,and gives simulations to verify the results.     

A Novel Floodless Service Discovery Mechanism Designed for Software-Defined Networking

The low-cost, self-configuration capability and ＂plug-and-play＂ feature of Ethernet establishes its dominant position in the local area networks （LAN）. However, it is hard to extend to large scale because of the legacy broadcast-based service discovery mechanism. Therefore, to solve this problem, a new split network architecture named Software-Defined Networking （SDN） is introduced in this paper, and a novel floodless service discovery mechanism （FSDM） for SDN is designed. For the FSDM, the widespread broadcast messages for Dynamic Host Configuration Protocol （DHCP） and Address Resolution Protocol （ARP） are considered especially, respectively. Then the DHCP relay and ARP proxy are proposed to handle DHCP broadcast messages and ARP broadcast messages, respectively. The proposed FSDM in this paper can eliminate flooding completely, reserve the auto- configuration characteristics. Particularly, there is no need to change the existing hardware, software and protocols of hosts for the proposed scheme. Finally, the simulation results are demonstrated to show that our proposed model allows redundant links existed in network and has the property of scalability, which can significantly reduce network traffic in data plane and control traffic in control plane, and decrease the overhead of control plane.     

Design and Implementation of a Novel HomePlug-Based Solution for Low Cost and High Performance Smart Home Networking

As the smart home is the end-point power consumer, it is the major part to be controlled in a smart micro grid. There are so many challenges for implementing a smart home system in which the most important ones are the cost and simplicity of the implementation method. It is clear that the major share of the total cost is referred to the internal controlling system network; although there are too many methods proposed but still there is not any satisfying method at the consumers’ point of view. In this paper, a novel solution for this demand is proposed, which not only minimizes the implementation cost, but also provides a high level of reliability and simplicity of operation; feasibility, extendibility, and flexibility are other leading properties of the design.     

Reliability Evaluation of All-User Terminals in LEO Satellite Communication Network Based on Modular Reduction

LEO satellite communication network has a large number of satellites distributed in low orbits,which leads to multiple coverage of many areas on the ground.It is hard work to describe and evaluate the reliability of LEO satellite communication network.To solve this problem,the reliability of all-user terminals in LEO satellite communication network is defined,and the corresponding reliability evaluation method is proposed in the paper.Due to the large scale of the interstellar network,a modular reduction algorithm using the modular network instead of the original network for state decomposition is proposed in this paper.Case study shows that the calculation time of the proposed method is equivalent to 6.28%of the original state space decomposition algorithm.On this basis,the reliability of LEO satellite communication network is further analyzed.It is found that the reliability of LEO satellite network was more sensitive to the reliability of Inter-Satellite link and the satisfaction of global coverage in the early stage,and it is more sensitive to the reliability of the satellite in the later stage.The satellite-ground link has a relatively constant impact on of LEO satellite network.     

Visual focus of attention estimation based on improved hybrid incremental dynamic Bayesian network

In this paper, a visual focus of attention(VFOA) detection method based on the improved hybrid incremental dynamic Bayesian network(IHIDBN) constructed with the fusion of head, gaze and prediction sub-models is proposed aiming at solving the problem of the complexity and uncertainty in dynamic scenes. Firstly, gaze detection sub-model is improved based on the traditional human eye model to enhance the recognition rate and robustness for different subjects which are detected. Secondly, the related sub-models are described, and conditional probability is used to establish regression models respectively. Also an incremental learning method is used to dynamically update the parameters to improve adaptability of this model. The method has been evaluated on two public datasets and daily exper iments. The results show that the method proposed in this paper can effectively estimate VFOA from user, and it is robust to the free deflection of the head and distance change.     

Utility optimization scheduling for multi-point video surveillance in ubiquitous network

Resource allocation is an important problem in ubiquitous network. Most of the existing resource allocation methods considering only wireless networks are not suitable for the ubiquitous network environment, and they will harm the interest of individual users with instable resource requirements. This paper considers the multi-point video surveillance scenarios in a complex network environment with both wired and wireless networks. We introduce the utility estimated by the total costs of an individual network user. The problem is studied through mathematical modeling and we propose an improved problem-specific branch-and-cut algorithm to solve it. The algorithm follows the divide-and-conquer principle and fully considers the duality feature of network selection. The experiment is conducted by simulation through C and Lingo. And it shows that compared with a centralized random allocation scheme and a cost greed allocation scheme, the proposed scheme has better per- formance of reducing the total costs by 13.0% and 30.6% respectively for the user.     

FEW-NNN: A Fuzzy Entropy Weighted Natural Nearest Neighbor Method for Flow-Based Network Traffic Attack Detection

Attacks such as APT usually hide communication data in massive legitimate network traffic, and mining structurally complex and latent relationships among flow-based network traffic to detect attacks has become the focus of many initiatives. Effectively analyzing massive network security data with high dimensions for suspicious flow diagnosis is a huge challenge. In addition, the uneven distribution of network traffic does not fully reflect the differences of class sample features, resulting in the low accuracy of attack detection. To solve these problems, a novel approach called the fuzzy entropy weighted natural nearest neighbor(FEW-NNN) method is proposed to enhance the accuracy and efficiency of flowbased network traffic attack detection. First, the FEW-NNN method uses the Fisher score and deep graph feature learning algorithm to remove unimportant features and reduce the data dimension. Then, according to the proposed natural nearest neighbor searching algorithm(NNN_Searching), the density of data points, each class center and the smallest enclosing sphere radius are determined correspondingly. Finally, a fuzzy entropy weighted KNN classification method based on affinity is proposed, which mainly includes the following three steps: 1、 the feature weights of samples are calculated based on fuzzy entropy values, 2、 the fuzzy memberships of samples are determined based on affinity among samples, and 3、 K-neighbors are selected according to the class-conditional weighted Euclidean distance, the fuzzy membership value of the testing sample is calculated based on the membership of k-neighbors, and then all testing samples are classified according to the fuzzy membership value of the samples belonging to each class;that is, the attack type is determined. The method has been applied to the problem of attack detection and validated based on the famous KDD99 and CICIDS-2017 datasets. From the experimental results shown in this paper, it is observed that the FEW-NNN method improves the accuracy and efficiency of flow-based network traffic attack detection.     

Facial expression recognition based on improved ResNet

Facial expression recognition(FER) is a vital application of image processing technology. In this paper, a FER model based on the residual network is proposed. The proposed model introduces the idea of the DenseNet, in which the outputs of the residual blocks are not simply added but are linked to the channel dimension. In addition, transfer learning is used to reduce training costs and accelerate training speed. The accuracy and robustness of the proposed FER model were tested by K-fold cross-v...     

指纹图像细化算法的研究

指纹识别是一种重要的生物特征鉴别技术。随着计算机技术的不断发展,自动指纹识别系统得到广泛的应用。因此进一步提高指纹识别的性能具有十分重要的意义,而指纹图像增强在指纹图像预处理过程中非常重要,直接影响指纹识别的识别率和识别速度。对指纹图像的细化算法进行了较深入的研究,分析了OPTA算法并且在OPTA算法的基础上,重新构建了细化模板,提出了一种新的细化算法.经过实验证明,该算法能够很好地满足细化的要求,细化完全彻底,细化以后的指纹骨架在纹线中心线,并保持了纹线原有的拓扑结构和细节特征,而且光滑无毛刺,运算速度也很快。     

异号权重模板CNN的似细化特性及其应用

本文发现并以定理的形式证明了具有异号权重模板的细胞神经网络系统在非均匀增益分段性输函数下的细 化稳态性性。     

干涉条纹中心线提取与细化的新方法

干涉条纹分析是光学干涉计量的重要分析方法,本文提出基于改进Yangtagai求极值点法与改进Hilditch细化法相结合的干涉条纹中心线提取与细化方法,该方法具有条纹断点少、抗噪声能力强以及处理速度快的特点。通过实验证实了本文提出方法的有效性。     

细化算法在手写体字符识别中的应用

在手写体字符识别的研究中,在提取特征的过程中需要提取字符的骨架信息,所以要对每一个字符进行细化处理。在研究ZS和LW细化算法的基础上,采用串行细化与并行细化相结合的方法,并引入一些模板改进算法使其能够有效去除多余枝杈和冗余像素的问题;通过对手写体数字、字母的文字进行测试,表明该细化效果较好,并有一定的实用价值。     

文字识别中的图象自适应预处理

为了提高文字识别的正确率，本文比较详细地探讨了手写文字图象在归一化、细化过程中存在的复杂的随机变形，提出了一些有特色的自适应性较强的预处理方法来解决诸如手写文字图象的总尺寸与其构成的笔划宽度一般不成正比等难题，我们还研究出比较符合人眼视觉效果的快速自适应细化算法（375字／秒），这些处理的结果能很稳定地保持原始文字的结构信息，为识别手写的汉字、字母、数字带来了方便。     

SDB-SOI晶片减薄技术综述

阐述了SDB-SOI晶片的减薄技术的特点和要求,依次介绍了化学机械抛光、电化学自停止腐蚀、等离子抛光技术和智能剥离技术的原理和特点,并指出了SDB-SOI晶片减薄的发展趋势。     

一种基于标记的并行细化算法

本文提出了一种基于图像标记的并行细化算法.首先计算图像中像素和轮廓距离的标记信息,并将标记图像作为细化算法的输入.通过综合利用标记信息,保留了几何特性,提高了细化算法的鲁棒性,同时为进一步图元识别/拟合奠定了基础.     

背照式CCD图像传感器减薄技术研究

提出了一种研磨减薄和化学腐蚀结合的方法,减薄CCD衬底。研磨减薄采用三氧化二铝的颗粒,减薄硅片至100 μm,化学腐蚀采用HF酸、硝酸、冰乙酸组成的混合溶液。实验表明,HF酸、硝酸、冰乙酸比例为2:1:11时,混合溶液对衬底与外延层的腐蚀速率比达到89:1。使用本技术减薄1 024×512可见光CCD,实现了背照式成像。     

硅片减薄技术研究

集成电路芯片不断向高密度、高性能和轻薄短小方向发展,为满足IC封装要求,越来越多的薄芯片将会出现在封装中。此外薄芯片可以提高器件在散热、机械等方面的性能,降低功率器件的电阻。因此,硅片减薄的地位越来越重要。文章简要介绍了减薄的几种方法,并对两种不同研磨减薄技术的优缺点进行了对比。此外,从影响减薄质量的因素如主轴转速、研磨速度及所使用的保护膜等几方面进行了实验的验证,分析了不同参数对质量的影响效果。并根据减薄后的质量情况,使用统计方法,对减薄的过程进行了监控。     

基于距离场细化的骨架提取算法

骨架保存了要处理对象的拓扑信息,是图像分析的重要研究内容之一.传统的骨架细化算法不能保证结果的准确性,而距离场的方法无法保证结果的连续性.为此提出一种快速有效的骨架提取算法,将经典的距离变换法和细化方法结合,克服二者之间存在的缺陷,实现算法的互补.经过大量实验验证,此方法能够得到连续、准确的骨架,可以很好地满足实际应用的需求.