Parametric FE-approach to flip-chip reliability under various loading conditions

In this paper we focus on the thermo-mechanical reliability of flip-chip assemblies which are, in addition to periodic thermal loads, constrained by mechanical boundary conditions caused by the attachment of a heat-spreader. Whereas mechanically unconstrained flip-chip assemblies have been in the focus of reliability studies for a long time, the loading induced by additional mechanical constraints and hence the impact on solder bump reliability is still largely unknown. So a comprehensive study was carried out comprising FE-simulations for lifetime prediction and thermal cycling tests for experimental verification. For this purpose a tool for modular parametric FE-model generation was developed. The experiments do coincide with the simulative prediction with good accuracy, allowing for the first time a distinct statement about the reliability of flip-chip packages with attached heat-spreaders. As a result we have found that in general all additional constraints on the chip do reduce its lifetime. A distinct ranking has be obtained as a function of the specified variables. Eventually design guidelines are given.     

基于非齐次泊松过程的深空骨干节点流量分析

在深空探测及通信中,由于通信距离及复杂深空环境的影响,使得通信难度较大,通信质量难以得到保证,组建深空通信网络是解决相关问题的有效途径。而深空骨干网作为深空通信网络的重要组成部分,其组成节点承担了较为繁重的数据转发服务请求,因此,对骨干网节点进行业务量分布分析是保证整个深空通信网络服务质量(QoS)及通信能力的前提。在假设服务请求到达率不为常数的情况下分析了深空骨干网节点的服务请求到达特点,得到了服务请求时间间隔低于限定时间的概率及某时间段内请求数量少于设定阈值的概率。为深空节点业务流量合理分配提供了一定依据,为深空通信数据路由选择提供了指导,即可避免节点拥塞,提升节点QoS提供能力,为深空通信网络路由及通信质量保证提供了较为直观的参考依据。     

深空通信网络协议的发展与展望

分析了发展深空通信网络的需求,结合深空通信网络的特点,对深空通信网络协议体系和路由策略等的技术现状进行了综述,对其中的关键技术进行了分析,并对深空通信网络的未来发展进行了展望。     

End-to-end encrypted network traffic classification method based on deep learning

Network traffic classification, which matches network traffic for a specific class of different granularities, plays a vital role in the domain of network administration and cyber security. With the rapid development of network communication techniques, more and more network applications adopt encryption techniques during communication, which brings significant challenges to traditional network traffic classification methods. On the one hand, traditional methods mainly depend on matching features on the application layer of the ISO/OSI reference model, which leads to the failure of classifying encrypted traffic. On the other hand, machine learning-based methods require human-made features from network traffic data by human experts, which renders it difficult for them to deal with complex network protocols. In this paper, the convolution attention network (CAT) is proposed to overcom those difficulties. As an end-to-end model, CAT takes raw data as input and returns classification results automatically, with engineering by human experts. In CAT, firstly, the importance of different bytes with an attention mechanism of network traffic is achieved. Then, convolution neural network (CNN) is used to learn features automatically and feed the output into a softmax function to get classification results. It enables CAT to learn enough information from network traffic data and ensure the classified accuracy. Extensive experiments on the public encrypted network traffic dataset ISCX2016 demonstrate the effectiveness of the proposed model.     

Performance evaluation of FDDI under asymmetric traffic conditions

FDDI is a proposed standard for optical fibre LANs. A simulation model is used to examine the performance of the FDDI protocol under asymmetric traffic conditions. The relationship between the average token rotation time and the relative stability of the network is also presented.<>     

深空测控通信网技术的发展与展望

介绍了美国喷气推进实验室(JPT)的概况及其在人类深空探测活动中所处的重要地位,进而介绍了深空测控通信网的地理分布、组成部分及使用设备,研究了美国深空网在信号覆盖、遥控能力、遥测能力及轨道测量方面的技术特点,最后对深空测控通信网技术的发展方向进行了归纳、分析和展望,提出了建立激光统一测控体制的构想.     

Distributed dynamic channel allocation algorithms in microcells under light traffic loading

We consider distributed dynamic channel allocation (DCA) algorithms for wireless cellular networks. Two algorithm extremes are the baseline of our investigation: Theaggressive type algorithms, where channel assignments of ongoing calls may be reconfigured to make room for a new access request; and thetimid one, where reconfigurations are not permitted. While, generally, analyses of DCA algorithms appear very difficult, we show here that a light traffic analysis can be tractable. For that realm, we derive performance expressions showing significant reductions in access blocking probability when an aggressive, rather than the timid, access algorithms is used. The relative efficacy of various algorithms that we present can be analytically probed in this range. Certain aggressive algorithms are observed to attain the best performance possible therein. The advantage over timid is particularly substantial for planar cellular arrays. These results hint that there may be significant channel savings, associated with the type of aggressive algorithms described here, in the important, yet much more difficult to analyze, operating realm where the blocking, but not necessarily the loading, is low; and lead us to devise a family of hybrid DCA algorithms that are both stable and low blocking.     

基于机器学习和深度学习的网络流量分类研究

随着互联网技术的不断发展以及网络规模的不断扩大,应用的类别纷繁复杂,新型应用层出不穷。为了保障用户服务质量(QoS)并确保网络安全,准确快速的流量分类是运营商及网络管理者亟须解决的问题。首先给出网络流量分类的问题定义和性能指标;然后分别介绍基于机器学习和基于深度学习的流量分类方法,分析了这些方法的优缺点,并对现存问题进行阐述;接着围绕流量分类线上部署时会遇到的3个问题:数据集问题、新应用识别问题、部署开销问题对相关工作进行阐述与分析,并进一步探讨目前网络流量分类研究面临的挑战;最后对网络流量分类下一步的研究方向进行展望。     

HTTP malicious traffic detection method based on hybrid structure deep neural network

In response to the HTTP malicious traffic detection problem,a preprocessing method based on cutting mechanism and statistical association was proposed to perform statistical information correlation as well as normalization processing of traffic.Then,a hybrid neural network was proposed based on the combination of raw data and empirical feature engineering.It combined convolutional neural network (CNN) and multilayer perceptron (MLP) to process text and statistical information.The effect of the model was significantly improved compared with traditional machine learning algorithms (e.g.,SVM).The F₁value reached 99.38% and had a lower time complexity.At the same time,a data set consisting of more than 450 000 malicious traffic and more than 20 million non-malicious traffic was created.In addition,prototype system based on model was designed with detection precision of 98.1%~99.99% and recall rate of 97.2%~99.5%.The application is excellent in real network environment.     

基于深度卷积神经网络的交通流量预测数学模型设计

精确的交通流量预测是实现未来智能交通的关键技术。神经网络模型在该领域的预测方面具有一定的优势。因此,为了提高预测精度,设计一种基于深度卷积神经网络的交通流量预测数学模型。首先,对交通流量数据的预处理方法进行分析,然后结合特征训练过程和卷积神经网络构建深度神经网络结构,并给出深度神经网络的配置参数。利用美国明尼苏达大学UMD分校的交通流数据集进行仿真实验,结果表明,提出的模型可以对短时交通全局趋势进行预测,并具有较好的稳定性和预测精度。     

深空双波段红外动态场景仿真与目标分析

为了提高深空中红外探测器探测波段选择的效率以及反映目标真实的在轨状态,文中研究了空间目标动态红外图像的生成技术以及双波段红外图像的仿真。结合从STK获得的空间目标的运行轨迹和姿态数据以及从3DS MAX获得的目标模型数据提出了一种进行空间目标红外成像仿真的方法。首先分析了深空背景中目标的红外辐射特性以及目标在中波和长波两个波段中的辐射特性,为探测器波段参数的选择提供了依据;然后介绍了用节点网络法求解热平衡方程的方法并且简化了求解过程,提高了效率;最后阐述了空间目标红外成像系统的理论模型,实现了空间目标的红外可视化建模与仿真。仿真图像对新型系统的研制和红外图像处理算法的研究有重要的意义。     

DOCSIS performance analysis under high traffic conditions in the HFC networks

Hybrid Fiber Coaxial (HFC) networks provide a wide bandwidth and represent the solution of choice for many residential networks nowadays. The Data Over Cable Service Interface Specification (DOCSIS) protocol is an important standard for HFC networks and is supported by the majority of current vendors. This protocol uses the Truncated Binary Exponential Back-off (TBEB) algorithm to resolve collisions within the network by means of a back-off window. However, the performance of this algorithm tends to deteriorate when the network load is high. Consequently, the present study develops a novel mathematical model for the TBEB algorithm and then uses this model to identify the window size which yields the optimal system throughput and minimum delay time under high traffic conditions. The present numerical results confirm that the performance is improved when the window settings identified from the developed model are applied.     

Crystallization of amorphous hydrogenated silicon films deposited under various conditions

The possibility of using the magnetron-assisted silane decomposition technique for the deposition of a-Si:H films as the basic materials for the production of polysilicon is analyzed. It is shown how specific features of the film structure affect the crystallization process.     

Performance evaluation of routing algorithms under various network configuration parameters

Routing is a relevant issue for maintaining good performance and successfully operating in a network. Many types of routing algorithms have been proposed, such as shortest-path, centralized, distributed and flow-based. This article not only defines the relationship among performance indicators and network configuration parameters, but also presents a practical method for improving routing decisions. © 1997 John Wiley & Sons, Ltd.     

基于无线帧的物理层网络编码同步误差性能

物理层网络编码充分利用无线信道的广播特性,能够有效提高网络吞吐量和带宽利用率。但是物理层网络编码的性能受到载波相位、载波频率、时间、采样钟不同步的影响,因此分析同步误差对物理层网络编码性能影响具有重要的研究意义。基于单载波系统,针对使用BPSK调制的无线通信系统,从理论上分析了一定帧长下,载波相位、载波频率、时间及采样时钟同步误差对系统性能的影响。分析结果表明,在载波相位和时间不同步两种情形下,物理层网络编码性能不受帧长影响;而在载波频率和采样时钟不同步两种情形下,系统性能均随帧长的增加而有所下降,其中采样时钟不同步情形受帧长影响最大。     

Megamp post regulators-practical design considerations to allowoperation under extreme loading conditions

Practical design equations and guidelines are established to allow magnetic amplifier (magamp) operation under extreme loading conditions. Three conditions are considered: shutdown of output, foldback of the output current, and operation under discontinuous inductor current mode. Design equations are derived for the three extreme operations, and their use is illustrated by numerical examples. The design equations provide the designers with tools for parameter trade-offs and lend themselves to computer-search design procedures     

多优化机制下深度神经网络的音频场景识别

随着并行计算能力的不断攀升和音频数据量的日益扩增,音频场景识别成为场景理解领域重要的研究内容之一。针对音频场景识别建模难度大和识别准确率不高的问题,本文提出了融合多优化机制的并行卷积循环神经网络算法模型。首先,将音频信号经预处理后转化为一定尺寸的梅尔声谱图,之后输入到网络模型中进行充分的空间特征和时间特征学习,最后进行识别。为了验证模型的有效性,在DCASE2019音频场景数据集上进行识别性能测试,结果显示,该算法模型对音频场景的识别准确率能够达到88.84%,优于传统网络模型,说明该算法模型对音频场景识别问题的有效性。      

Multiscale nature of network traffic

The complexity and richness of telecommunications traffic is such that one may despair to find any regularity or explanatory principles. Nonetheless, the discovery of scaling behavior in teletraffic has provided hope that parsimonious models can be found. The statistics of scaling behavior present many challenges, especially in nonstationary environments. In this article, we overview the state of the art in this area, focusing on the capabilities of the wavelet transform as a key tool for unraveling the mysteries of traffic statistics and dynamics