Forward-backward method for scattering from imperfect conductors

The previously developed forward-backward method for calculating scattering from perfectly conducting azimuthally homogeneous surfaces is extended to imperfect conductors, where the dielectric constant has a large imaginary part such as sea water at X-band (10 GHz). An example shows that highly accurate results at X-band are obtained for the case of a steepened sea wave     

Forward-backward iterative physical optics algorithm for computing the RCS of open-ended cavities

The forward-backward methodology is combined with the iterative physical optics (IPO) algorithm to improve convergence for cavity scattering problems. Wave propagation inside elongated cavities, such as jet engine inlet ducts, follows a predominant down-and-back path. The forward-backward method allows the IPO currents on the cavity walls to be updated sequentially (forward) and reverse-sequentially (backward) along the waveguide axis. A relaxation parameter is introduced to help control the convergence characteristics, making the new algorithm mathematically equivalent to the classical iterative method of symmetric successive over-relaxation. The fast far-field approximation (FaFFA) accelerates the matrix-vector products in the IPO formulation, and an equivalent surface impedance is used to characterize thin material linings in the cavity.     

Forward-backward method for scattering from dielectric rough surfaces

The iterative forward-backward (FB) method is a recently proposed efficient technique for numerical evaluation of scattering from perfectly conducting rough surfaces. Extension of the method to include scattering from imperfect conducting surfaces, with a high imaginary part of the complex dielectric constant, has also been proposed. The FB method is further generalized to analyze scattering from dielectric rough surfaces with arbitrary complex dielectric constant. Electric and magnetic equivalent surface currents are split into forward and backward components and equations governing these current components are obtained. As a solution, an iterative scheme is proposed and its convergence rate is analyzed. Finally, the effectiveness of the method is assessed by comparing the obtained scattering results with "exact" ones, computed by employing the usual method of moments (MoM).     

Forward-backward method with spectral acceleration for scattering from Layered rough surfaces

A fast method of moments is presented to calculate electromagnetic wave scattering from layered one-dimensional rough surfaces. The formulation is provided for M stratified homogeneous regions, separated by M-1 rough surfaces, and solved using point matching and pulse basis functions. Compared to the single surface case, the solution of scattering from M-1 surfaces requires significantly more memory and computational time. To facilitate the solution, the forward-backward method with spectral acceleration is applied. As an example, a dielectric layer on a perfect electric conductor surface is studied. The numerical results are compared with the analytical solution for layered flat surfaces to partly validate the formulation. The accuracy, efficiency, and convergence of the method are then studied for various rough surfaces and layer permittivities.     

A new method for computing nonlinear carrier diffusion insemiconductor optical devices

The solution of carrier diffusion equations typically associated with semiconductor optical devices has been achieved by combining a function expansion scheme, using the Hermite-Gauss functions as the basis set, with the collocation numerical procedure. Results for a wide range of cases obtained by this new scheme compare very favorably with those calculated with other methods. Not only is the present process computationally fast and efficient, but it has the added attraction of providing the basis for conveniently solving also the nonlinear electromagnetic wave equation for the self-consistent modeling of such devices     

采用多核处理器进行可重构并行仿真的方法

文中对多传感器视觉信息处理算法进行分析,根据可重构处理器的并行计算参数模型提出了一种并行计算仿真的方法。多核处理器环境中,每个线程在独立的核上运行,线程间具有并发性。利用并发的线程模拟可重构阵列单元（PE）的运算方式,调用OpenMP设置多个线程并行执行,在多核计算机平台上模拟可重构处理器的计算过程。利用此方法能在没有具体的PE连接方案前,通过使用计算核模拟PE单元,将算法映射到多核处理器环境中。通过分析算法在多核计算机上的并发执行效率,来优化视觉信息算法在可重构阵列上的映射方案。     

Ray analysis of low-grazing scattering from a breaking water wave

Low-grazing-angle backscattering from a modeled breaking-wave surface profile has been calculated using a ray-optical approach and compared with reference scattering found using an extended moment method. The calculations show that interference between the direct backscatter from the breaking plume and multipath scattering between the plume and wave face can lead to the HH-to-VV polarization-backscattering ratios of greater than 9 dB that characterize sea-spike events. The multipath effects can be accurately predicted from simple reflection from the front face at the smallest grazing angles. At higher angles, diffraction from rapid changes in the surface curvature must also be considered     

Real-time computing: a new discipline of computer science andengineering

This paper surveys the state of the art in real-time computing. It introduces basic concepts and identifies key issues in the design of real-time systems. Solutions proposed in literature for tackling these issues are also briefly discussed     

A new algorithm for computing the reliability of complex networks by the cut method

It is well known that computing network reliability is equivalent to computing network unreliability. A fusion of the reliability branching algorithm (RBA) with the path method has given a new method for symbolic reliability analysis. The present paper gives a new method for symbolic unreliability; it is a combination of the RBA with the cut method. The algorithm is simple for computing by hand and has easy execution for the use of computers.     

Efficient method for computing double matrix products

We introduce an efficient method for computing matrix products of the formY=AXB, whereA andB are sparse and constant. We analyze the complexity of the method, develop quantitative criteria for determining when it can be used effectively, and demonstrate its use in a Kalman filter.Part of this work was supported by the GE Aerospace Business Group while the authors were in residence at the GE Research and Development Center, Schenectady, New York 12301, USA.     

自适应计算为MPEG-4注入新的动力

下一代无线和便携式通信产品将广泛应用MPEG-4标准的流式视频和音频。相比基于DSP的设计,采用自适应算法能够更加有效地实现高质量流式视频所要求的强大处理能力。 比如,一个由编码通道和解码通道(全彩色、分辨率为352×288像素)所组成的双通道流式30fps MPEG-4视频信号要求的运算速度为1400000000次/秒。MPEG-4编码包括离散余弦变换(DCT)和运动估计等计算密集型算法(如图)。     

An entropy-based method for computing software structural complexity

A new method for computing software structural complexity based on the entropy evaluation of the random uniform response function associated with the so-called software characteristic function is proposed. The functions, parameters and algorithms to perform this computation are introduced and some applications are given. This complexity metric allows different designs to be compared and was used successfully in the assessment of safety critical redundant software.     

一种车联网的边缘计算构建方法

文章基于移动边缘计算相关研究,参考现网主要架构,针对车联网中大数据量和低时延要求提出了移动边缘计算服务器在现网中的部署位置和具体架构、数据传输流程,通过流量分流网关进行流量分流,然后经解包后把车联网数据发送至平台处理,反馈信息也由相同链路传送回终端.     

Humanistic computing: “WearComp” as a new framework andapplication for intelligent signal processing

Humanistic computing is proposed as a new signal processing framework in which the processing apparatus is inextricably intertwined with the natural capabilities of our human body and mind. The emphasis of this paper is on personal imaging applications of humanistic computing, to take a first step toward an intelligent wearable camera system that can allow one to effortlessly capture our day-to-day experiences, help us remember and see better, provide us with personal safety through crime reduction, and facilitate new forms of communication through collective connected humanistic computing. The author's wearable signal processing hardware currently provides the computational power of a UNIX workstation concealed within ordinary-looking eyeglasses and clothing. Thus it may be worn continuously during all facets of ordinary day-to-day living, so that, through long-term adaptation, it begins to function as a true extension of the mind and body     

Simple method of computing a function in fixed-point arithmetic

The authors present a simple method, based on the principle of successive approximation, of computing a function using a fixed-point arithmetic processor. As an example computing the square root function is considered and the performance is compared with Newton's method, in terms of accuracy and the number of instruction cycles required     

MobiSoC: a middleware for mobile social computing applications

Recently, we started to experience a shift from physical communities to virtual communities, which leads to missed social opportunities in our daily routine. For instance, we are not aware of neighbors with common interests or nearby events. Mobile social computing applications (MSCAs) promise to improve social connectivity in physical communities by leveraging information about people, social relationships, and places. This article presents MobiSoC, a middleware that enables MSCA development and provides a common platform for capturing, managing, and sharing the social state of physical communities. Additionally, it incorporates algorithms that discover previously unknown emergent geo-social patterns to augment this state. To demonstrate MobiSoC's feasibility, we implemented and tested on smart phones two MSCAs for location-based mobile social matching and place-based ad hoc social collaboration. Experimental results showed that MobiSoC can provide good response time for 1,000 users. We also demonstrated that an adaptive localization scheme and carefully chosen cryptographic methods can significantly reduce the resource consumption associated with the location engine and security on smart phones. A user study of the mobile social matching application proved that geo-social patterns can double the quality of social matches and that people are willing to share their location with MobiSoC in order to benefit from MSCAs.

On a method computing transient wave propagation in ionospheric regions

A consequence of an exoatmospheric nuclear burst is an electromagnetic pulse (EMP) radiated from it. In a region far enough away from the burst, where nonlinear effects can be ignored, the EMP can be represented by a large-amplitude narrow-time-width plane-wave pulse. If the ionosphere intervenes the origin and destination of the EMP, frequency dispersion can cause significant changes in the original pulse upon reception. A method of computing these dispersive effects of transient wave propagation is summarized. The method described is different from the standard transform techniques and provides physical insight into the transient wave process. The method, although exact, can be used in approximating the early-time transient response of an ionospheric region by a simple integration with only explicit knowledge of the electron density, electron collision frequency, and electron gyrofrequency required. As an illustration of the method, it is applied to a simple example and contrasted with the corresponding transform solution.     

一种新型高速四进制光计算算法研究

提出了一种面向多级调制格式的先进高速光计算算法,利用高非线性光纤中超快简并和非简并四波混频(FWM),采用100Gb/s正交相位调制(DQPSK)信号,仿真实验实现了50Gbaud/s高速光计算,包括加法、双定向减法、补码、加倍等光计算功能。当比特误码率(BER)为10-9时,测得高速光计算的功率代价小于4dB。     

A simple method for computing pyramidal horn gain proximity corrections

Pyramidal horn gain calibration by the two-antenna or three-antenna method, as well as measurements in the anechoic chamber that make use of standard gain horns, often require the application of proximity corrections. In this paper, a simple pyramidal horn proximity correction method is presented. This method is derivable directly from Jull's finite-range gain formulas, and is nearly as accurate as the frequently used and more laborious method of Chu and Semplak.