Extending the threshold and frequency range for phase-basedfrequency estimation

Standard phase-based frequency estimation has a threshold that is high, is frequency dependent, and does not decrease with increasing signal length. These problems are solved by processing the signal with a highly overlapped filter bank before applying phase-based frequency estimation. By exploiting decimation, a closed-form matrix inversion, and cascades of simple filter banks, the computational complexity is kept low     

动力学神经网络模型的构建及其稳定性研究

针对静态人工神经网络具有在反映系统动态行为时网络结构复杂、不能很好地反映系统动态性能的缺点,提出一种由带有积分器和可调反馈系数的神经元构成的新型动力学神经网络模型。该网络比以前的动态网络即递归网络或在此基础上改进的网络能更好地反映系统的动态性能,网络的结构更加简单,训练过程加快,从而使系统能够更好的运行。利用梯度下降法研究了该网络的权值调整算法,并通过李雅普诺夫稳定性判据讨论了这种新型动力学神经网络的稳定性条件。该网络研究为反映系统的动力学行为提供了更好的模型结构和理论算法,为神经网络的发展提供了新的研究方向。     

Efficient estimation of a time-varying dimension parameter and its application to EEG analysis

This paper considers the problem of estimating the dimension of nonstationary electroencephalogram (EEG) signals and describes the implementation of an efficient algorithm to calculate a time-varying dimension estimate. The algorithm allows the practical calculation of a dimension estimate and its statistical significance over large data sets with a high temporal resolution. The method is applied to EEG recordings from patients with temporal lobe epilepsy and in one case the results of the analysis are compared with those obtained from an existing method of computing the correlation density.     

Bayes shrinkage estimation of reliability and the parameters of a finite range failure time model

A finite range failure time distribution has been proposed and studied. For estimating the two parameters of this distribution, this paper considers a prior assumption that (1 − b) is the probability that the scale parameter θ and shape parameter p have the values θ₀ and p₀, respectively, and that the rest of the probability mass b(0 ≤ b ≤ 1) is distributed as h(p,θ) = h₁(p)h₂(θ). The value h₁(p) is a uniform density for p and h₂(θ) is an inverted gamma density for θ. With this prior density, Bayes estimators are first obtained and then Bayesian shrinkage estimators are defined. Bayesian shrinkage estimators are compared with maximum likelihood estimators (m.l.e.) and it was found that the proposed estimators are better than m.l.e. for quite a wide range of parameters.     

一种快速时变衰落信道估计方法研究

由于现代无线通信系统对快速移动通信服务的要求越来越高,传统的信道估计方法在快速时变信道环境下的估计性能下降幅度很大。文章提出了一种快速时变信道环境下的信道估计方法,在发射端发射2个调频率相反的LFM信号,在接收端在峰值搜索的过程中结合谱校正技术搜索信号幅度谱的峰值和对应点的坐标,实现谱峰值的超分辨率估计,利用分数阶傅里叶变换的时移和频移性质,根据发射端信号的峰值点与收端信号的峰值点之间的差值,估计出信道的时延和频移,并给出信道的衰落幅度和初始相位的估计方法。仿真结果表明,该方法较传统信道估计方法的估计精度有较大的提升且实时性高。     

Linear and nonlinear ARMA model parameter estimation using anartificial neural network

Addresses parametric system identification of linear and nonlinear dynamic systems by analysis of the input and output signals. Specifically, the authors investigate the relationship between estimation of the system using a feedforward neural network model and estimation of the system by use of linear and nonlinear autoregressive moving-average (ARMA) models. By utilizing a neural network model incorporating a polynomial activation function, the authors show the equivalence of the artificial neural network to the linear and nonlinear ARMA models. They compare the parameterization of the estimated system using the neural network and ARMA approaches by utilizing data generated by means of computer simulations. Specifically, the authors show that the parameters of a simulated ARMA system can be obtained from the neural network analysis of the simulated data or by conventional least squares ARMA analysis. The feasibility of applying neural networks with polynomial activation functions to the analysis of experimental data is explored by application to measurements of heart rate (HR) and instantaneous lung volume (ILV) fluctuations     

An FPGA based scalable architecture of a stochastic state point process filter (SSPPF) to track the nonlinear dynamics underlying neural spiking

Recent studies have verified the efficiency of stochastic state point process filter (SSPPF) in coefficients tracking in the modeling of the mammalian nervous system. In this study, a hardware architecture of SSPPF is both designed and implemented on a field-programmable gate array (FPGA). It provides a time-efficient method to investigate the nonlinear neural dynamics through coefficients tracking of a generalized Laguerre–Volterra model describing the spike train transformations of different brain sub-regions. The proposed architecture is able to process matrices and vectors with arbitrary sizes. It is designed to be scalable in parallel degree and to provide different customizable levels of parallelism, by exploring the intrinsic parallelism of the FPGA. Multiple architectures with different degrees of parallelism are explored. This design maintains numerical precision and the proposed parallel architectures for coefficients estimation are also much more power efficient.     

Global stability of a class of neural networks with time-varying delay

This paper presents a new sufficient condition for the uniqueness and global asymptotic stability of the equilibrium point for a class of neural networks with time-varying delays. The result is obtained by the use of a more general type of Lyapunov-Krasovskii functional, establishing a relation between the network parameters of the neural system and time-varying delay parameter. The result is also shown to be a generalization of a previously published result.     

基于小波神经网络和ARMA模型的CPI波动区间预测

目前对于消费者价格指数(CPI)的预测研究基本集中于点预测。为预测本期较上期的CPI数据波动区间,提出一种基于小波神经网络和ARMA组合模型预测的方法。该模型首先利用小波神经网络对CPI数据进行拟合测试,对测试序列实际输出和期望输出的残差序列{et}进行ARMA建模预测,然后基于方差最小原则得到预测残差序列{e!t}95%的置信区间。通过实验表明预测残差序列95%置信区间可以很好的反应未来CPI数据的波动情况,具有较高的参考价值。     

Comparison of DDE and ETDGE for time-varying delay estimation

The performances of DDE and ETDGE, which are two recently proposed methods for the direct estimation of time delay between signals received at two spatially separated sensors, are compared. Although both algorithms are computationally efficient, it is shown that the ETDGE method generally outperforms the DDE method for tracking nonstationary delays with different source signals     

一维动态海面的电磁散射杂波模拟和参数估计

通过计算机模拟不同风速下随时间变化的动态海面，将传统的两种用于计算静态粗糙面的方法(前后向选代法和基尔霍夫近似)结合起来，计算了动态海面电磁散射回波，从而获得了不同海况、不同极化状态和不同入射角(包括掠入射)条件下海杂波的统计描述，并通过计算几种典型杂波模型的“似然率”，对所得数据进行了最佳拟合。计算结果与文献中的数据吻合。     

Semi-empirical model for the threshold voltage of a double implanted MOSFET and its temperature dependence

A simple and accurate semi-empirical model for the threshold voltage of a small geometry double implanted enhancement type MOSFET, especially useful in a circuit simulation program like SPICE, has been developed. The effect of short channel length and narrow width on the threshold voltage has been taken into account through a geometrical approximation, which involves parameters whose values can be determined from the curve fitting experimental data. A model for the temperature dependence of the threshold voltage for the implanted devices has also been presented. The temperature coefficient of the threshold voltage was found to change with decreasing channel length and width. Experimental results from various device sizes, both short and narrow, show very good agreement with the model. The model has been implemented in SPICE as a part of the complete d.c. model.     

Identification of time-varying biological systems from ensembledata [joint dynamics application]

The theory underlying a new method for the identification of time-varying systems is described. The method uses singular value decomposition to obtain least-squares estimates of time-varying impulse response functions from an ensemble of input-output realizations. No a priori assumptions regarding the system structure or form of the time-variation are required and there are few restrictions on the input signal. Simulation studies, using a model of time-varying joint dynamics, show that the method can track rapid changes in system dynamics accurately and is robust in the presence of output noise. An application of the method is demonstrated by using it to track dynamic ankle stiffness during a rapid, voluntary, isometric contraction. During the transient phase of the contraction, low-frequency ankle stiffness gain decreased in a manner which could not be described with the second-order model of joint dynamics often used under stationary conditions.     

Autonomic cardiovascular control following transient arousal from sleep: a time-varying closed-loop model

Recent studies suggest that exposure to repetitive episodes of hypoxia and transient arousal can lead to increased risk for cardiovascular disease in patients with obstructive sleep apnea syndrome (OSAS). To obtain an improved understanding of and to quantitatively characterize the autonomic effects of arousal from sleep, a time-varying closed-loop model was used to determine the interrelationships among respiration, heart rate and blood pressure in 8 normal adults. A recursive least squares algorithm was used in combination with the Laguerre expansion technique to estimate the time-varying impulse responses of the 4 model components. We found that during arousal: 1) respiratory-cardiac coupling gain increases in nonrapid-eye movement (NREM) but not in REM sleep; 2) in both NREM and REM sleep, baroreflex gain shows an initial increase, but this is followed by a more sustained decrease below pre-arousal baseline levels, allowing sympathetic tone to be elevated over a relatively long duration; 3) the gains of other model components show increases with arousal that are consistent with the increased sympathetic modulation of systemic vascular resistance and contractility of the heart. These findings establish a normative database against which further measurements of cardiovascular arousal responses in OSAS may be compared.     

Deterministic regression methods for unbiased estimation of time-varying autoregressive parameters from noisy observations

A great deal of interest has been paid to autoregressive parameter estimation in the noise-free case or when the observation data are disturbed by random noise. Tracking time-varying autoregressive (TVAR) parameters has been also discussed, but few papers deal with this issue when there is an additive zero-mean white Gaussian measurement noise. In this paper, one considers deterministic regression methods (or evolutive methods) where the TVAR parameters are assumed to be weighted combinations of basis functions. However, the additive white measurement noise leads to a weight-estimation bias when standard least squares methods are used. Therefore, we propose two alternative blind off-line methods that allow both the variance of the additive noise and the weights to be estimated. The first one is based on the errors-in-variable issue whereas the second consists in viewing the estimation issue as a generalized eigenvalue problem. A comparative study with other existing methods confirms the effectiveness of the proposed methods.     

Optimal range and Doppler centroid estimation for a ScanSAR system

This paper presents a new range and Doppler centroid estimation algorithm for a ScanSAR system. This algorithm is based on processing the image data in the overlapped region of two bursts of the same beam or adjacent beams. It leads to highly accurate radar pointing angles that are paramount to achieving good radiometric performance in ScanSAR imagery. The achievable accuracy is derived theoretically and verified by tests performed using SIR-C ScanSAR data and ERS data. This algorithm is computationally efficient and easy to implement. The proposed Doppler centroid estimation algorithm is also an excellent candidate for a strip mode SAR system     

Research and application of a neural network classifier based on dynamic threshold

In this study, a Multi-Layer BP neural network (MLBP) with dynamic thresholds is em-ployed to build a classifier model. As to the design of the neural network structure, theoretical guidance and plentiful experiments are combined to optimize the hidden layers’ parameters which include the number of hidden layers and their node numbers. The classifier with dynamic thresholds is used to standardize the output for the first time, and it improves the robustness of the model to a high level. Finally, the classifier is applied to forecast box office revenue of a movie before its theatrical release. The comparison results with the MLP method show that the MLBP classifier model achieves more satis-factory results, and it is more reliable and effective to solve the problem.     

Application of neural algorithms for a real-time estimation of ozone profiles from GOME measurements

The thermal structure of trace gases, their distribution in the atmosphere, and their circulation mechanisms result from a complex interplay between radiative, physical, and dynamical processes. Neural-network algorithms can be a useful tool to face such complexities in retrieval operations. In this paper, their potentialities have been exploited to design real-time procedures for the estimation of vertical profiles of ozone concentration from spectral radiances measured by GOME, the first instrument of the European Space Agency capable of monitoring global distribution of ozone and other trace gases.     

Reflection of electromagnetic wave from a time-varying medium

For the purpose of radiowave propagation in the ionosphere, in a plasma and in other media, the reflection and transmission of an electromagnetic wave from a suddenly changed conductivity, when a plane wave is normally incident upon an interface separating two conducting media, are studied. The field expressions are exactly obtained through the use of the Laplace transform. The discussion is primarily on the time behaviour of the reflected wave.     

Global output convergence of a class of continuous-time recurrent neural networks with time-varying thresholds

This paper discusses the global output convergence of a class of continuous-time recurrent neural networks (RNNs) with globally Lipschitz continuous and monotone nondecreasing activation functions and locally Lipschitz continuous time-varying thresholds. We establish one sufficient condition to guarantee the global output convergence of this class of neural networks. The present result does not require symmetry in the connection weight matrix. The convergence result is useful in the design of recurrent neural networks with time-varying thresholds.