未来数据网络需求分析

网络是信息交换的中枢机构,网络的技术创新是应用和业务创新的基础和使能条件。应用基于网络,网络要基于对未来人类智能社会的假设和愿景进行顶层设计。首先介绍了未来典型应用对数据网络提出的新需求,然后分析了当前数据网络存在的技术差距,最后尝试定义了未来8~10年的数据网络的新能力,旨在为未来数据网络的设计提供指引。     

Aircraft interface with the future ATC system

The authors outline how the advantages of improved navigational accuracy can benefit both the individual aircraft and the air transportation system despite a mixed environment. The development of advanced navigation, flight planning, performance optimization, automatic guidance, and other capabilities in current and future flight management systems, together with the availability of digital communications between the airplane and ATC, suggests the need to review the potential airplane contributions to the ATC system of the future. A number of application areas utilizing current airborne flight management technology have been proposed for use with ATC that could provide significant operational benefits. These include: time-navigation applications in traffic management; multisensor navigation to provide automatic dependent surveillance; improved precision for approach, missed approach, and departure procedures; and improved navigational accuracy supporting reduced airway widths, holding pattern airspace, and aircraft separation standards. Several examples of operational benefits for operators and ATC during the transition period are identified, and the need for integrating these features into future ATC capabilities and procedures is stressed     

Optical requirements for projection lithography

An image modulation criterion is proposed as a simple optical performance parameter for lithography using positive photoresist. The modulation, or contrast, is defined in terms of the maximum and minimum intensity in the image of a line/space pattern as (I_MAX ? I_MIN)/(I_MAX + I_MIN). For lithography directly on integrated circuit wafers, a contrast in the range of 80–90% gives acceptable linewidth control. Universal curves are derived showing image contrast for diffraction-limited optical systems with partially coherent illumination and including defocus error. The allowable defocus is derived as a function of the contrast requirement and pattern linewidth. Based on these optical considerations, and the properties of positive photoresists, the useful resolution of state-of-the-art projection systems are shown to be in the neighborhood of 1.25 μm, and progress to a useful resolution of about 0.5 μm is forseeable.     

Workforce considerations: maintenance of the future ATC system

The authors describe the development of a remote maintenance monitoring system, the rationale for consolidation of the field maintenance workforce, and the development of an automated modeling tool, the work center consolidation model. The model is described, and the methodology used to validate a node substitution algorithm is explained in detail. The node substitution algorithm has been implemented and incorporated into the work center location model (WCLM). The improved WCLM has the combined benefits of the dBASE III command language and the reduced processing requirements of the node substitution algorithm. This provides user friendliness and fast and flexible data manipulation, while the node substitution algorithm allows the WCLM to model regionwide problems in reasonable time and with reasonable effort. The improved WCLM has been successfully used to model the US Federal Aviation Administration (FAA's) Southern Region. The WCLM was able to model the various configurations required by the region, and it produced the optimal combination of work centers acceptable to the region     

Electrode flatness requirements for cathode projection microfabrication systems

One of the most promising techniques for large-area submicron electron-beam lithography is the Cathode Projection Microfabrication System. According to published reports, these systems suffer from excessive image distortion. This distortion is partly caused by nonflat photocathodes and silicon wafer anodes. In this paper, a universal curve is obtained which quantitatively predicts and explains the image distortions which result from surface topography. Apparently, the practicality of multilayer submicron lithography in such systems is presently threatened by a requirement that nonreproducible wafer surface waviness be less than 2-3 microns.     

Signal processing applications of oblique projection operators

Oblique projection operators are used to project measurements onto a low-rank subspace along a direction that is oblique to the subspace. They may be used to enhance signals while nulling interferences. In the paper, the authors give several basic results for oblique projections, including formulas for constructing oblique projections with desired range and null space. They analyze the algebra and geometry of oblique projections in order to understand their properties. They then show how oblique projections can be used to separate signals from structured noise (such as impulse noise), damped or undamped interfering sinusoids (such as power line interference), and narrow-band noise. In some of the problems addressed, the oblique projection provides an alternative way to implement an already known solution. Expressing these solutions as oblique projections brings geometrical insight to the study of the solution. The geometry of oblique projections enables one to compute performance in terms of angles between signal and noise subspaces. As a special case of removing impulse noise, the authors can use oblique projections to interpolate missing data samples. In array processing, oblique projections can be used to simultaneously steer beams and nulls. In communications, oblique projections can be used to remove intersymbol interference     

未来网络场景及需求分析综述

面对现今网络在可扩展性、移动性、服务质量保障、管理性等方面的严峻挑战,通过对未来网络应用场景的分析,归纳总结了未来网络的关键能力需求,提出了未来网络的主要功能和性能指标。最后,对未来网络的发展演化趋势进行了展望。     

A look into the future of data acquisition

The basic Fabry-Perot interferometer sensor is made of a cavity formed by a gap in a fiber optic cable. The end of the fiber is minimally reflecting. Losses associated with a simple cleavage of the fiber eliminate the need for treating the remaining fiber as a cavity     

Freeway-traffic data processing

Two basic data processing problems associated with freeway traffic are formulated: estimation of traffic variables (section mean speed and density) and detection of occurrence of an incident or accident within a given section of the road. The existing traffic sensors and the current estimation techniques and their shortcomings are reviewed. Stochastic models for traffic variables are developed and used to design optimum recursive estimators. The incident detection problem is introduced as a potential area where many interesting research topics exist.     

A statistical technique for processing radio interferometer data

The output of a radio interferometer is the Fourier transform of the object under investigation. Due to the limited coverage of the Fourier plane the reconstruction of the image of the source is blurred by the beam of the synthesized array. In this paper a maximum-likelihood processing technique is described that uses the statistical properties of the received noise-like signals. This technique has been used extensively in the processing of large aperture seismic arrays. This inversion method results in a synthesized beam that is more uniform, has lower sidelobes, and higher resolution than the normal Fourier transform methods. The maximum-likelihood method (MLM) algorithm was applied successfully to very long baseline (VLB) and short baseline interferometric data.     

国务院第一次全国经济普查领导小组办公室数据处理组副组长倪志良谈--第一次全国经济普查数据处理的安全

全国第一次经济普查将于2004年12月31日在全国展开。这次经济普查与以往搞过的人口普查、工业普查、基本单位普查等单项普查不同,是一项大规模的综合性普查。这次普查的对象是中国境内从事第二、第三产业活动的全部法人单位、产业活动单位和个体经营户,其规模之大、范围之广、内容之丰富在我国普查史上是前所未有的,也是有史以来调查样本最多、数     

未来航空电子高速数据总线技术的研究

高速数据总线技术是航空电子综合化最重要的支撑技术之一,直接决定整机航电系统综合化程度的高低,本文结合美军JAST计划,分析了高速数据总线技术的若干问题,针对JAST计划中统一网络的首选协议SCI做了实际的网络性能测试,给出了结果。     

Display requirements for future man—Machine systems

The primary purpose of this paper is to explore the parameters of display devices that might better serve the needs of highly interactive man-machine systems. Examples of what is meant by "highly interactive" are those systems designed to aid editors in the acquisition, correction, and production editing of "camera-ready" textual information, assist with the development of design drawings in computer-aided engineering design, and provide for development and management information systems graphic reports that can be interacted with by the user. CRTs are the most versatile and ubiquitous form of display in the machine-to-man communication part of this interface. Two major technical drawbacks to CRTs for this function are the size of the work space and the resolution. To significantly increase either of these greatly increases the costs that are already too high for many applications. For the man-to-machine portion of the interface, the CRT has some undesirable characteristics.     

一种新型的用于高速串行接口的数据处理电路

本文提出了一种新型的用于USB2.0高速模式下(480Mbps的数据传送率)的数据处理电路.这种电路采用八位并行的方法将数据转换为USB协议规定的数据格式,包含一个高速、低功耗的并串转换电路及一个八分频电路.芯片设计基于TSMC公司的0.25μmCMOS混合信号模型,采用半定制(semi-custom)的设计流程.电路的前后仿真结果表明该数据处理电路达到了480MHz的传输速度,符合USB2.0的要求.     

增量式正交局部判别投影法

针对现有投影分析算法随着输入数据量的增加计算复杂度急剧增长这一问题,通过子块优化策略构建了目标投影模型,称之为增量的局部判别投影(ILDP)算法。算法兼顾样本的类间离散度和类内紧凑性,求得的投影矩阵还具有正交性;通过子块叠加和奇异值升级算法对模型的求解进行了增量式扩展,计算过程中并无出现矩阵逆操作,即规避了小样本问题。在COIL图像库、USPS手写字体库和ExYaleB人脸库中的实验表明,对比经典的ILDA、LSDA、MMP等降维算法,ILDP具有更高的识别率,尤其在USPS数据库中,ILDP的识别率接近于90%,而其它的算法识别率都低于85%。与此同时,ILDP的计算量也明显少于对比算法,在USPS数据库中仅需要少于0.5s的时间即可完成最优投影矩阵计算。     

Progressive quantized projection approach to data hiding.

A new image data-hiding technique is proposed. The proposed approach modifies blocks of the image after projecting them onto certain directions. By quantizing the projected blocks to even and odd values, one can represent the hidden information properly. The proposed algorithm performs the modification progressively to ensure successful data extraction without the need for the original image at the receiver side. Two techniques are also presented for correcting scaling and rotation attacks. The first approach is an exhaustive search in nature, which is based on a training sequence that is inserted as part of the hidden information. The second approach uses wavelet maxima as image semantics for rotation and scaling estimation. Both algorithms have proved to be effective in correcting rotation and scaling distortion.     

基于数据挖掘的快速记录存储器数据处理技术

针对目前国内缺少专门分析快速记录存储器(QAR)数据的有效手段的情况,研究了一种新的基于数据挖掘的QAR数据的分析方法。首先结合聚类和概率分析对k-means算法进行改进,解决了聚类数目难以确定的难题,形成了良好的聚类效果;然后,在此基础上结合加权最小距离分类器及概率分析的方法,对待分类的QAR数据的类别属性进行判断以确定异常数据;最后给出了仿真实验,验证了该方法的可行性和有效性。