Intemet路由器级拓扑测量中目标选择方法研究

为提高Intemet路由器级拓扑测量的完全性。给出了完备目标集合的定义，介绍了已有测量目标选择方法并分析其不足之处，提出了“以桩网络而不是IP地址为抽样粒度”的观点和“桩网络完全覆盖，两阶段随机选择”的完备测量目标选择方法。实验结果表明，从测量目标的角度来说。这种测量目标选择方法保证了测量的完全性。     

动态的时间驱动的分组采样技术

提出了一种用于网络测量的动态时间驱动分组采样技术,根据网络流量的变化动态调整采样次数,从而能将最后的测量误差率控制在一个给定的范围内,在保证精度的同时极大地降低了测量的代价。并且该方法的实现复杂度与“1out of N”采样技术近似,文章中给出了有关的理论分析和实验研究。实验结果表明该方法达到了设计的目标。根据网络运行状态动态调整网络测量方案的思想可以广泛用于网络测量当中。     

基于卷积神经网络的SAR图像目标检测算法

该文研究了训练样本不足的情况下利用卷积神经网络(Convolutional Neural Network, CNN)对合成孔径雷达(SAR)图像实现目标检测的问题。利用已有的完备数据集来辅助场景复杂且训练样本不足的数据集进行检测。首先用已有的完备数据集训练得到CNN分类模型,用于对候选区域提取网络和目标检测网络做参数初始化;然后利用完备数据集对训练数据集做扩充;最后通过四步训练法得到候选区域提取模型和目标检测模型。实测数据的实验结果证明,所提方法在SAR图像目标检测中可以获得较好的检测效果。     

基于移动互联网模式的4G客户性能感知预判技术研究及应用

本项目通过终端侧感知指标采集打通终端到网络的最后1 km的瓶颈,结合网络侧多样完备的信令,“网络”、“终端”联动建立完备的客户感知模型,自主开发感知问题溯源定界分析算法,通过即时探测分析的流计算技术,实现大规模的客户感知预判,先于用户发现问题,先于投诉解决问题。     

网络测量系统研究中亟待解决的若干问题

网络很难测量．不断扩大的网络和日益繁多的测量需求使有效的网络测量变得更加困难。实现灵活高效的网络测量系统和方法一直是网络测量研究中最重要的目标之一。首先，本讨论了网络测量的目标和应用．综述了当前网络测量系统和方法的研究工作。然后，在此基础上。提出了网络测量系统中亟待解决的若干重要问题：网络测量系统的同一性结构、多测量工具的动态协作、可扩展性、测量信息的建模和综合、自动化监测方法、网络测量的执行特性分析、资源授权和分配机制、最优化测量任务调度、指标量化问题。最后是本的总结和针对这些问题正在开展的研究工作。     

“网络测量与行为学”课程教学探讨

互联网的发展使得网络测量和行为分析成为监控、理解、认识网络进而实现对网络精细化管理的重要方法。我院于2009年开始为研究生开设了“网络测量与行为学”课程。本文探讨了该课程设置的必要性和教学方法,并给出该课程所涉及的相关资料,以期吸引国内高校能够开设该课程,进而培养更多的网络测量与行为学相关人才。     

宽带中国战略下的同轴网发展前景

2013年8月17日,国务院发布了“宽带中国”战略实施方案,部署未来8年宽带发展目标及路径,意味着“宽带战略”从部门行动上升为国家战略。本文梳理了广电网络发展的三个阶段,分析了国内双向网改情况,认为广电完全可以依靠同轴网络来实现三网融合的宽带中国战略。     

基于单站测量的纯仰角目标参数航迹滤波方法

根据特殊战场环境下传感器测量目标的不完全信息的情况,提出了空间目标纯仰角参数的处理思想,并给出了一种基于投影关系的参数航迹滤波方法,该方法基于斯特瓦尔特定理和总体最小二乘法,讨论了等高飞行目标的纯仰角处理形式和斜直线目标的纯仰角参数滤波形式。仿真验证了该方法能较好适应特殊战场环境下的不完全测量信息情况。     

多跳无线网网络拓扑控制优化目标研究

研究了多跳无线网中网络性能与网络拓扑结构关系的问题。数值计算结果表明网络节点只与“最邻近”的节点建立链路时形成的拓扑结构节点的吞吐量最大,这样的拓扑结构同时能满足网络的低功耗、低截获概率和可靠性要求。计算机仿真结果验证了上述结论。因而,网络拓扑控制的目标就转化为节点在邻节点中选择“最邻近”节点,继而建立逻辑链路的问题。     

多基地雷达的目标定位

针对多基地雷达可提供多种测量参数这一特点,本文提出了一种利用最小二乘法则来估计目标位置的方法,并探讨了提高定位精度的可能性。为理论分析方便,文中引入了“误差矢量”的概念,从理论上探讨了目标参数对定位精度的影响及测量参数的选择等问题。通过对多普勒效应的简单讨论,得出了位置信息与速度信息之间的对应关系,这种对应关系使文中的分析结论有了更广泛的意义。     

Cooperative Wireless Medium Access Exploiting Multi-Beam Adaptive Arrays and Relay Selection

Cooperative transmission among wireless network nodes can be exploited to resolve collisions and thereby enhance the network throughput. Incorporation of multi-beam adaptive array (MBAA) at a base station/access point (destination) receiver has been shown to improve the network performance. In this paper, we propose an efficient cooperative wireless medium access scheme that exploits novel relay selection methods in a network equipped with MBAA at the destination receiver. Unlike existing techniques that require the estimation of angles-of-arrival (AoAs), the proposed scheme uses the spatial correlation among users for simpler but more effective collision detection and resolution. We present two useful relay selection methods based on channel gain and spatial correlation. It is shown that the joint use of an effective relay selection method and an MBAA in a wireless network can significantly improve the uplink throughput. The throughput of the proposed scheme and its upper bound are analytically derived. Numerical and simulation results have demonstrated significant performance enhancement achieved by the proposed cooperative wireless medium access scheme.     

容延容断网络路由算法及其存在问题的探析

容延容断网络(delay/disruption tolerant network,简称DTN)是一种新型网络体系结构.在这一类网络中,节点与节点之间不存在端到端的路径,因此节点之间的通信一般使用"存储-携带-转发"的模式来进行,也就是节点需要先将信息存储起来再携带信息,直到遇到能够将消息转发到目的节点的时候才将消息转发出去.因此在容延容断网络中路由算法的关键性问题是如何以较高的发送成功率和较低的延迟将消息转发到目的节点.DTN网络的应用场景通常由手机、蓝牙设备以及具备WiFi功能的手持电子设备等组成的移动自组网络.本文主要研究DTN路由算法及其存在的问题.     

A hardened network-on-chip design using runtime hardware Trojan mitigation methods

Due to the globalized semiconductor business model, malicious hardware modifications, known as hardware Trojans (HTs), have risen up as a big concern for chip security. HT detection and mitigation methods for general integrated circuits have been investigated in the past decade. However, the majority of the existing efforts are not customized for HTs in Networks-on-Chip (NoCs). To complement the firmware and software level methods for rogue NoCs detection, we propose countermeasures to harden the NoC hardware design against tampering. More specifically, we propose a collaborative dynamic permutation and flit integrity check method to mitigate the potential inside-router HTs inserted by the disloyal member in the NoC design house or the 3^rd-party system integration company. Our method improves the number of received packets by up to 70.1% over the other methods if the HT controls the NoC packet destination address. The average link availability of our method is 43.7% higher than that of the exiting methods. Our method increases the effective average latency by up to 63.4%, 68.2%, and 98.9% for the single HT in the destination, header, and tail fields, respectively, over the existing methods.     

基于直觉模糊推理的威胁评估方法

该文针对威胁评估问题,提出一种基于直觉模糊推理的评估方法。首先,分析了联合防空作战中空天来袭目标影响威胁评估的主要因素、威胁评估的不确定性,以及现有威胁评估方法的特点与局限性,建立了威胁程度等级划分的量化模型。其次,设计了系统状态变量的属性函数,讨论了在模糊化策略方面对输入变量进行的量化和量程变换方法。再次,建立了系统推理规则,设计了推理算法和清晰化算法,分析了规则库中所包含规则的完备性、互作用性和相容性,给出了规则库的检验方法。最后,以20批典型目标的威胁评估实例,验证了方法的有效性。     

协同DF-OFDM系统中基于MRC的机会中继研究

针对机会中继译码转发协同系统中未考虑直传链路的问题,提出一种由目的节点根据系统中断概率进行中继选择的方案,研究了将机会中继信号与直传信号在目的节点进行最大比合并时的系统性能,推导了采用最大比合并时的系统中断概率。研究结果表明,当直传链路存在时,通过目的节点的中继选择,可使系统性能得到进一步提高,其性能优于不考虑直传链路时的机会中继系统。     

Joint selection of adaptive relaying mode in cooperative transmission

Relay is a promising technology in wireless communications. There are several relaying modes for different channel conditions. In this article, an adaptive relaying mode selection scheme by destination and relay is proposed. The proposed scheme changes the signal forwarding mode at the relay station. It adaptively chooses among amplify-and-forward (AF), decode-and-forward (DF) and direct mode according to cyclic redundancy check (CRC) detection at relay and destination. Moreover, block error rate (BLER) and throughput are adopted to evaluate system performance. The simulation results demonstrate that the proposed scheme significantly outperforms other existing relaying mode selection schemes.     

基于直觉模糊推理的态势预测方法

在作战辅助决策系统中,针对战场态势预测问题,提出一种基于直觉模糊推理的战场态势综合预测方法.分析了当前态势预测方法的特点与不足;建立了直觉模糊推理规则,设计了推理合成算法和解模糊算法,检验了推理规则的完备性,并引入态势指标的权重向量;最后通过态势预测中的一个具体仿真实例验证了该方法的有效性与合理性.     

An overview of electrical characterization techniques and theory for IC packages and interconnects

This paper reviews current approaches to the electrical characterization and modeling of IC packages and interconnects. An overview of both frequency and time-domain measurement methods and summaries of equivalent circuit model selection and extraction methodologies are included. Additionally, an overview of numerical methods for electromagnetic modeling is included for completeness. Finally, relevant case studies from the literature are summarized to further supplement the discussed techniques. The focus is primarily on high-frequency signal related characterization and power integrity issues are not directly considered in this paper. This paper is presented in the context of the growing requirement for package and interconnection electrical models for high-speed and miniaturized systems.     

Estimating the Traffic Distribution Matrix of a Packet-Switched Network Using Aggregate Traffic Measurements

Methods are developed to estimate the source–destination traffic distribution matrix of a packet network using only aggregate link and source/sink traffic measurements. The methods are useful for network planning and monitoring of large packet networks, where it is impractical to gather measurement data on every end-to-end traffic flow. The first method assumes that the distribution matrix is time-invariant. This method is of theoretical interest but provides the foundation for developing a method for the realistic case of a time-varying matrix. The second method assumes that the matrix is time-varying. It uses linear programming (LP) to find a distribution matrix that optimally fits the measurement data. A practical problem with the first two methods is that the computational requirements increase as the square of the number of network nodes. The third method is a fast exact decomposition procedure for the time-invariant case that scales with the network size. The maximum number of unknowns that needs to be solved simultaneously is equal to the number of network nodes. The final method is a fast decomposition procedure for the time-varying case. This procedure scales with the network size. It uses LP to find an approximate distribution matrix that optimally fits the measurement data. The methods are applied to simulated example networks to illustrate the accuracy and speed.