Fault simulation using small fault samples

This article emphasizes simulation-based sampling techniques for estimating fault coverage that use small fault samples. Although random testing is considered to be the primary area of application of the technique it is also suitable for estimating the fault coverage of nonrandom tests based on specific fault models. Especially for fault coverages exceeding 95%, it is shown that a precise estimate can be obtained using a fault sample of only 500 faults. The estimation is based on a binomial approximation of the probability density of the sample fault coverage. Using Bayes statistics an estimate is obtained whose accuracy is a linear function of the sample size if the fault coverage approaches 100%. The sample size is independent of the circuit size, thus making fault sampling particularly interesting for the fault simulation of ULSI designs due to the resulting reduction of the time complexity of fault simulation from O(N ₂) to O(N).This work was performed while Dr. Daehn was with the Laboratorium fuer Informationstechnologie at the university of Han- nover, Germany.     

Detection of tone signals using a nonlinear energy tracker

A simple scheme is proposed for detecting tone signals narrowly spaced in frequency. By employing Teager's energy tracker as a preprocessor, the tone detector can be implemented without the use of sharp notch filters. The proposed detector is applied to distinguishing ITU-T V.8 and V.25 tone signals. The detector performance is analysed and verified by computer simulation     

Consistent estimation of continuous-time signals from nonlinear transformations of noisy samples

A signal cannot in general be reconstructed from its sign, i.e., from its hard-limited version. However, by the deliberate addition of noise to samples of the signal prior to hard limiting, it is shown that the signal can be estimated consistently from its hard-limited noisy samples as the sampling rate tends to infinity. In fact, such estimates are shown to converge with probability one to the signal and to be asymptotically normal. Although the estimates are in general nonlinear, they can be made linear by a proper choice of the noise distribution. These rather unexpected results hold for all bounded and uniformly continuous signals. In addition to the hard-limiter, such results are also established for certain monotonic and nonmonotonic nonlinearities.     

利用非线性自适应技术检测微弱的瞬时信号

本文根据局部流形理论,提出了一类非线性自适应预测滤波模型。仿真结果显示,本文提出的自适应预测滤波模型具有很强的非线性近似能力,可以检测淹没在超混沌或杂波环境中的瞬时信号。     

采用尺度空间理论的红外弱小目标检测方法

为了检测红外场景中尺寸大小变化的弱小目标,针对传统滤波方法中固定大小滤波核对此类特性目标检测表现出的不足,提出一种基于尺度空间理论的红外弱小目标检测方法。首先对弱小目标特性进行分析,提出采用点扩散函数形式的目标模型来描述弱小目标;采用固定自适应邻域的方法对原始红外图像进行预处理,抑制背景杂波,增强目标能量;依据尺度规范化后的拉普拉斯尺度空间对图像不同元素滤波响应的不同,获取图像中的可疑目标,利用可疑目标点与其周围像素的梯度关系得到可疑目标点的中心坐标,并据此得到其在图中的尺寸大小;对每个可疑目标划分一个自适应大小窗口,获取分割阈值,分割出真实目标。实验结果表明,该方法能较好地检测出弱小目标,且具有较低的虚警率。     

Stable identification of nonlinear systems using neural networks: theory and experiments

This paper presents an approach for stable identification of multivariable nonlinear system dynamics using a multilayer feedforward neural network. Unlike most of the previous neural network identifiers, the proposed identifier is based on a nonlinear-in-parameters neural network (NLPNN). Therefore, it is applicable to systems with higher degrees of nonlinearities. Both parallel and series-parallel models are used with no a priori knowledge about the system dynamics. The method can be considered both as an online identifier that can be used as a basis for designing a neural network controller as well as an offline learning scheme for monitoring the system states. A novel approach is proposed for the weight updating mechanism based on the modification of the backpropagation (BP) algorithm. The stability of the overall system is shown using Lyapunov's direct method. To demonstrate the performance of the proposed algorithm, an experimental setup consisting of a three-link macro-micro manipulator (M/sup 3/) is considered. The proposed approach is applied to identify the dynamics of the experimental robot. Experimental and simulation results are given to show the effectiveness of the proposed learning scheme.     

Combating highly imbalanced steganalysis with small training samples using feature selection,

We consider a particular paradigm of steganalysis, namely, highly imbalanced steganalysis with small training samples, in which the cover images always significantly outnumber the stego ones. Researchers have rigorously studied sampling and learning algorithms as well as feature selection approaches to the class imbalance problem, but the research in the steganalysis domain is rare. This study provides a systematic comparison of eight feature selection metrics and of three types of methods developed for the imbalanced data classification problem in the steganalysis domain. Each metric is compared across three different classifiers and four steganalytic features. The efficiency of the metrics is evaluated to determine which performs best with minimal features selected. The performance of the three types of methods and their combinations is examined. Moreover, we also investigate the effect of feature dimensionality, sample number and imbalance degree on the performance of feature selection inresolving imbalanced image steganalysis.     

Recovery of signals from nonuniform samples using iterative methods

An iterative method to recover a bandlimited signal from its ideal nonuniform samples is proposed. The convergence of iterations is proved, and general regions for convergence are found. It is shown that the iterative method is also applicable to other forms of nonuniform sampling, i.e. natural sampling and interpolated sampling (such as sample-and-hold signal). Simulation results show that this method works effectively and fairly fast, and the errors after a few iterations are negligible if a particular sufficient condition is satisfied or the sampling rate is higher than the Nyquist rate     

Parametric reliability-prediction based on small samples

Reliability prediction is considered for systems that fail due to graceful degradation of operational efficiency to below an acceptable level. The efficiency is represented by a stochastic process, X(t), either scalar or vector. The system reliability with regard to such failures is parametric. It is measured by the probability, R(τ), that X(t) is within the acceptable limits during the time period, [0,τ]. To predict parametric reliability, a procedure is proposed to determine the lower confidence bound at confidence-level, q, for R(τ). Measurement results for small N, e.g., 3-5, of X(t) at times, 0⩽t₁⩽t₂⩽…i ⩽τ are taken as initial data along with stated assumptions on the nature of X(t) variations. The small-sample feasibility is due to the prior selection of the failure-model (on the basis of the system-analogue observation) while the test results are used to estimate the model parameters     

采用D-S证据理论的双色红外小目标融合检测

针对远距离低信噪比条件下目标检测难的实际问题,提出采用D-S证据理论的双色红外小目标融合检测方法。该方法首先采用统计检测方法对各传感器图像进行目标检测处理;接着采用“或”逻辑对各传感器的目标检测结果进行融合,以降低目标漏检的可能性;然后在各传感器图像中提取融合检测结果中各候选目标区域的多个图像特征作为进一步消除虚警的证据;最后采用D-S证据理论对各候选目标区进行基于多特征的目标融合识别处理,并将识别结果作为整个系统最终的目标检测输出。实验结果证明了该算法的有效性。     

Piecewise linear implementation of nonlinear dynamical systems: from theory to practice

It is shown how to implement by a mixed-signal circuit a continuous-time dynamical system. The chosen case study is the Hindmarsh-Rose model of a biological neuron, but the design strategy can be applied to a large class of continuous-time nonlinear dynamical systems. The system nonlinearities are first approximated by using piecewise-linear functions and then digitally implemented on a field programmable gate array. The linear part of the system is completely analogue and is implemented by using operational amplifiers. Measurement results show that the circuit can reproduce the main dynamics of a biologically plausible neuron.     

The nearest-neighbor rule for small samples drawn from uniform distributions (Corresp.)

It is shown in the classification problem, when independent samples are taken from uniform distributions, that for small sample sizes the probability of misclassification when using the nearest-neighbor rule is "close" to its asymptotic value. It is also shown that when using this rule the probability of classification in many cases is close to its Bayes optimum even for small sample sizes. Moreover, if one is restricted to a small sample size from one population, it is shown that it is not necessary to "make up" this deficiency by taking a large sample from the other population; best results may be obtained when both sample sizes are small.     

Circuit theory of periodically driven nonlinear systems

The systems under discussion are nonlinear, but driven by a strong periodic "carrier." Examples of such systems are oscillators, frequency multipliers, limiters, modulators, and periodically driven feedback systems. It is of interest to inquire how small perturbations on the periodic driving are propagated throughout the system, and to this end a circuit theory for these perturbations is developed. In different Contexts these perturbations could be desired or undesired modulation, noise, hum, or synchronizing signals. In general the random processes in such a system will, because of the periodic driving, be nonstationary; but various representations are developed that are stationary, and hence can be described by spectral analysis. The concept of impedance is developed for the small perturbations, and the validity of Kirchhoff's Laws is examined. Specific problems are not treated in detail; instead, a general framework is set up within which a variety of problems can be analyzed. Problems of this sort include the theory of noise in oscillators, propagation of noise and modulation in nonlinear systems, the noise theory of frequency multipliers, and synchronization of oscillators.     

Detection of epileptical seizures based on alpha band statistical features

Wireless Personal Communications - Research in the field of bistatic radars have growing interest with exclusive importance in defence sector, aerospace industry, remote sensing, meteorological and...     

Some aspects of the theory of nonlinear networks

In the development of network theory over the years, the primary focus of attention has been in the area of linear systems. Several reasons for this emphasis can easily be cited, but perhaps the foremost reason is that it has long been thought that, except in certain very special cases, little progress toward a rigorous definitive theory could be expected once the hypothesis of linearity is discarded. The recent success in the use of numerical methods for computing solutions of the equations for specific nonlinear networks (the importance of which is not to be minimized) has, furthermore, resulted in a certain complacency on the part of many engineers who occasionally need to solve network problems. One senses their outlook as being, basically, that whenever a particular nonlinear problem arises, one need only then run, data in hand, to the computer. Somewhat ironically, however, the development of computer-aided network analysis techniques has also been a prime impetus for many of the recent theoretical investigations in the field of nonlinear networks, and although much remains to be done, a rather comprehensive body of knowledge in this area has begun to take form. A number of related recent contributions to the theory of non-linear networks are reviewed here. As distinct from the computational aspects of the network analysis problem, we discuss work whose primary purpose is to yield an understanding of the nature of the equations that describe the behavior of nonlinear networks, and to identify and relate certain properties of the network elements, and the manner of their interconnection, to properties of the equations and their solutions. In addition, we do frequently touch on the problem of computation since, as has already been implied, it is indeed one of the purposes of the work discussed here to provide more of a theoretical foundation on which to base the numerical analyses.     

Some problems of asymptotic theory of nonlinear waves

Asymptotic methods developed recently in the nonlinear wave theory and applicable to the analysis of nonsinusoidal processes are discussed. These methods are based on the assumption of sufficiently slow space-time modulation of wave parameters (shortwave approximation). The intention of this paper is neither to review all problems of asymptotic wave theory nor to give a complete bibliography. Its purpose is to show the main ideas of the methods and to present concrete procedures for the construction of the approximate solutions that correspond to the "nonlinear geometrical optics" for various types of waves. A brief classification of these processes is also given in this paper. The following questions are considered: a) the asymptotic method for quasi-stationary waves; b) a modification of this method describing slowly varying aperiodic waves (shocks and solitary waves); c) the derivation of "averaged variational principle" and its generalization for dissipative systems; d) the method of reduction of a system of equations close to the linear hyperbolic ones which is valid for nonstationary waves propagating along the characteristics; and e) the concrete problem of cylindrical finite-amplitude electromagnetic wave propagation in isotropic dielectrics.     

Efficient randomized Web-cache replacement schemes using samples from past eviction times

The problem of document replacement in Web caches has received much attention and it has been shown that the eviction rule "replace the least recently used document" performs poorly in Web caches. Instead, it has been shown that using a combination of several criteria, such as the recentness and frequency of use, the size and the cost of fetching a document, leads to a sizable improvement in hit rate and latency reduction. However, in order to implement these novel schemes, one needs to maintain complicated data structures. We propose randomized algorithms for approximating any existing Web-cache replacement scheme and thereby avoid the need for any data structures. At document-replacement times, the randomized algorithm samples N documents from the cache and replaces the least useful document from the sample, where usefulness is determined according to the criteria mentioned above. The next M相似文献