一种新的为实现IP Over WDM设计的光波长路由器

将IP与WDM相结合是一个极具吸引力的研究方向。与以往点到点的方式不同,提出了一种新的实现IP over WDM的方案,称为DPDP（缺省波长和专用波长,default path and dedicated path）,给出了详细的节点结构图,并解释了其工作流程。最后对相关问题进行了讨论。     

基于前缀区间和前缀层次(PRPL)的IPv6路由查找算法

通过对前缀区间（PR）和前缀层次（PL）的研究,并结合IPv6全球单播地址和IPv6骨干网路由表的特征,提出了一种基于前缀区间和前缀层次（PRPL）的高性能的IPv6路由查找算法。通过使用合成的不同数据量的IPv6路由表对该算法进行了测试,测试结果表明该算法的可扩展性非常出色,它能够支持数据量达到。200k的大路由表。与同样环境下的经典算法的测试结果进行了比较,比较结果也表明该算法在平均查找速度和内存消耗方面都有一定的优势。出色的性能和对大路由表的支持,使得该算法能够应用到骨干网IPv6路由器中。     

动态快速路由查找算法

给出了动态快速路由查找算法（DFR）的原理和实现。该算法采用特殊的数据结构来构建索引表,能支持动态插入、删除和更新路由;DFR算法最多4次访存,最少2次访存,就能找到完整的路由信息。该算法不仅适用于软件实现,而且由于查找简单,也适合于硬件实现。     

一种用于MPLS光波长路由器的自适应路由策略

针对现有MPIS光波长路由器存在IP层和光层分别路由的不足,提出了一种使用了联合路由算法的自适应路由策略,该路由策略可以适应流量疏导端口资源的变化,合理利用波长和流量疏导端口,提高网络性能和资源利用率。同时为了支持该路由策略,还提出了一种辅助拓扑的构造方法。仿真结果显示,该自适应路由策略具有更好的性能,可以承载更多的业务。     

基于MEMLS的查找表法雪深反演

利用分层积雪微波辐射模型(MEMLS)和积雪先验知识建立了新疆地区冬季雪深与特定微波频率通道亮温差(BTD)之间的查找表,并针对不同的地形以及坡向因素对积雪参数进行了修正以提高MEMLS输出精度.然后利用先进微波扫描辐计2(AMSR2)被动微波遥感数据反演了2013年1月和2014年1月整个新疆地区的每日积雪深度信息,并使用气象台站的观测数据以及实地测量数据进行精度验证.结果表明,该模型对雪深的估算精度要优于常用的Chang模型,可较好地用于反演新疆地区的积雪深度信息.     

一种新的SeaWinds散射计海面风场反演方法

针对传统的最大似然法存在的主要缺点,在分析SeaWinds散射计模型函数表的基础上,根据雷达后向散射截面积（σ^0）随风速和相对方位角的变化规律,提出了一种新的海面风场反演方法。然后利用SeaWinds散射计部分L2A实测数据对该方法进行了验证．并在等风向间隔的条件下将该方法与传统的最大似然方法对同一实验数据的运行时间进行了比较。实验结果表明,与传统方法相比,该方法运行效率有明显提高,且反演结果与美国L2B反演结果十分接近。     

关于在交换局域网进行主动捕包的研究

对在交换型以太网络下的捕包方法进行研究,提出了利用底层协议伪装攻击结合IP转发技术来进行主动捕包的模糊,分析了主动捕包可以利用的几种底层协议伪装攻击及其特点。     

一种新型的基于数据平面共享环的IP over ASTN机制

提出了一种重叠模型下新型的基于数据平面共享环的IP over ASTN机制,其优点是利用IP网络中路由器自身具备的能力和在ASTN边缘节点引入基于以太网的二层交换能力,以较低的环路带宽代价实现各IP网间可达性信息传达,同时避免了对路由器的基本运行机制的较复杂修改。分析了相应的组网形式、边缘节点结构和网络运行机制,并对其中具有关键意义的带宽限制参数提出两种参数确定算法。     

数字视频后处理IP模块的VLSI设计

在讨论数字视频后处理增强算法的基础上,提出了一种数字视频信号后处理IP模块的VLSI实现方法。针对视频信号处理实时性以及运算量大的特点,提出了基于同步并行流水线的VLSI结构。应用0．35μm CMOS工艺对该IP模块设计进行了综合验证,并成功地将其应用于数字化处理电视的视频格式转换专用芯片VPP860的设计中。     

协作系统中DMPSK查表调制的译码方案研究

采用一种新的查表差分算法,理论分析得出,此差分算法不仅能降低硬件实现难度,而且在调制端避免了传统差分的数学运算.随后又在检测端针对多符号差分检测(MSDD)复杂度高的问题,提出结合自更新半径的球形译码(Auto-Updating Radius Sphere Detection,AUR-SD)算法的检测策略.仿真表明,结合了此两种算法的方案不仅使得复杂度得到了明显的改善,也使得在性能上相比传统宽度优先球形译码检测算法更加逼近ML算法.因此,可以作为一种较好的方案应用在差分协作系统中.     

A high speed 2-bit correlator chip for radio astronomy

The design and performance of a 16-channel 2×2-b CMOS correlator chip with 22-b on-chip integration, operating at a typical clockrate of 64 MHz, is presented. The chip, developed by the Netherlands Foundation for Research in Astronomy (NFRA), is used for the construction of large wideband correlation spectrometers. A 1024-channel module is described as an application example     

高速三体船研究综述

从水动力理论、模型试验和船型优化等方面对高速三体船的研究进展进行了综述，阐述了高速三体船的应用前景，归纳评述了有关的研究成果和尚存在的问题，并提出了关于这种新船型的若干研究方向。     

大规模灾害条件下基于连通支配集的重路由方案

针对在大规模灾害条件下,网络拓扑变化剧烈,现有的洪泛机制将会恶化网络性能的情况,提出了一种基于连通支配集的重路由方案,能够在拓扑发生剧烈变化后,快速提供恢复路径来确保节点间重要业务的快速可达.该方案应用连通支配集划分网络.通过只允许连通支配集内的路由器转发链路状态信息,来限制产生的链路状态信息的副本数量,以减小网络拓扑的剧烈变化对网络性能的冲击,并根据不同拓扑变化提供连通支配集的快速更新、重计算策略.仿真结果表明,该方案能够有效降低网络中链路状态信息报文的数量,同时不增加网络收敛时间.     

A high speed data acquisition and analysis system for ultrasonicenergy measurements

A high-speed microcomputer-based data acquisition and analysis system was designed and constructed. The system was interfaced to an ultrasound sector scanner. The primary application of the system, in this case, was to measure the power spectrum of the backscattered (low-level) ultrasound signal from acoustic inhomogeneous media, such as liver tissue. The system was bench-tested and then tested successfully in normal operative conditions. The system digitizes the ultrasound signal, with sample rate of 16 Msample/s and stores the digitized data in a temporary storage (static random-access memory). When the temporary storage is full, the data is transferred to a microcomputer, where a fast Fourier transform algorithm is applied to calculate the power spectrum of the ultrasound signal. Pattern-recognition algorithms are applied to differentiate between normal and abnormal liver tissues using the acoustic attenuation coefficient as the differentiation feature     

A high speed drive coil for a large capacity bubble memory

We calculated an eddy current loss in a wire as a function of an applied ac field. Estimating the field at an actual coil wire and summing the losses in all points of wires we obtained a formula representing a coil loss. Based on the calculated results, we designed coils which drive thirty-two 64 kb chips. The observed loss of the coils was in good accord with the calculated one. This 2 Mb module was operated at 40 Oe rotation field of 500 kHz with the coil loss of 8 W.     

IP核心网中VoIP业务服务质量的初步研究

通过对比传统电话网的语音业务特点对IP核心网中的VoIP业务流进行了模拟,并通过仿真对IP核心网中VoIP业务流特性进行了观察分析。结果发现,大量VoIP业务汇聚产生的业务流具有突发性小、较平稳的流量特性,因而不必为其分配过多的带宽资源;研究了Internet自相似业务流特性对VoIP业务性能的影响,提出了在VoIP业务中采用高优先调度机制等技术来保障VoIP服务质量的网络设计原则。     

A non–stroboscopic system for examining high speed rotating objects

A non–stroboscopic system for freezing high speed rotational motion has been developed by WaddelI. The system depends on the fact that the image of a stationary object (e.g., turbine disc, crankshaft, grinding wheel) can be made to rotate accurately about one axis only (i.e., no oscillation which will blur the image) when the image is passed through a certain prism. The image of the object will rotate twice for one revolution of the prism. If the object now rotates, and a fixed velocity ratio is maintained between the object and the prism, the rotating image of the object is rendered optically stationary by the prism. The image is rendered stationary through 360 degrees of object rotation. A novel and inexpensive synchronising system has been evolved by Waddell to maintain a stationary image of the object for long periods and over a large range of speeds and to maintain that static image despite transients in the speed of the object.     

A new method for the electro-optical measuring of fast high voltage pulses

This report describes a new method for measuring fast high voltage electric pulses based on one particular application of the Kerr effect. This work was carried out in order to further the development of optical particle detectors. At the present time with our method, the minimum risetime of the measured pulse reaches the nanosecond domain while the maximum voltage is limited only by the physical dimension of the electro-optical transducer.The measuring principle chosen requires a new calibration method in order to reconstruct the observed signal.The use of nitrobenzol as the active liquid allows the continuous modification of the cell sensitivity over a ratio of 1 to 3 without changing the geometry of the cell.     

Precracking high speed steel for fracture toughness testing

Fracture toughness testing of high speed steel, which has a high fatigue strength and low fracture toughness, is a problem because fatigue cracks are difficult, or impossible, to initiate at a maximum fatigue stress intensity of $\text{0.7 K}{}_{\mathrm{<mtext>IC</mtext>}}$, as specified. A method of initiation by the use of an electric pen and subsequent fast propagation by fatigue has been studied and a procedure developed to give accurate, reproducible values of $\text{K}{}_{\mathrm{<mtext>IC</mtext>}}$ on subsequent fracture toughness testing.     

基于振动的发动机转速高精度动态测量新方法

提出一种无需编码盘和鉴相装置的往复发动机瞬态转速测量新方法.该方法是基于发动机振动信号最低谐次频率获得发动机的估计转速,在发动机转速平稳或波动不大情况下采用离散频谱校正技术对估计转速进行校正以提高测量精度;在发动机处于升速或降速等非平稳状态时则利用获得的估计转速对发动机进行阶次跟踪,利用阶次谱获得估计转速校正参数,从而实现对发动机瞬态转速的动态测量.台架实验和不同缸数的实车测试都验证了该算法在发动机转速提取上具有很高的精度.较之在文献中常见的基于编码器的转速测量方法,这种方法硬件简单,特别适用于在用车日常维护和检测中使用.