摘要:In the deterministic inversion approaches for aperture synthesis radiometers (ASRs) imaging,the statistical property about the prior of brightness temperature distribution (BTD) is not taken into account.Aimed at the extended sources with discontinuous BTD,a sparse prior based statistical inversion approach (SIA) for ASRs imaging is proposed.According to the SIA,a modified difference operator is used to extract the implicit sparse prior about the discontinuous BTD of the extended sources,and the equivalent hierarchical prior Gaussian model for the sparse prior probability distribution is constructed.Then the inversion of ASRs is recast as hyperparameter estimation,and the expectation maximization algorithm is proposed to estimate the hyperparameter.The simulations and experiments show that the sparse prior based SIA can improve the radiometric accuracy of the reconstructed image and is more robust to the impacts of the imperfections of the ASRs as compared to the deterministic inversion approaches.
摘要:This paper mainly studies the extended self-similarity of sea clutter frequency spectrum and the application of multi-scale Hurst exponent to target detection within sea clutter.As a generalization of the fractional Brownian motion,the extended self-similar process uses the multi-scale Hurst exponent to describe fractal signals.The multi-scale Hurst exponent can characterize the details of fractal signals in different scales,which makes up for the deficiency of the mono-Hurst exponent that can only describe the whole roughness of fractal signals.Based on real radar data,this paper first studies the extended self-similarity of real sea clutter frequency spectrum and the influencing parameters.Then,the characteristic that the multi-scale Hurst exponent in the optimal frequency scale is relatively sensitive to the target is utilized for designing CFAR detection algorithm for target detection within sea clutter.The analytic results of real data show that the multi-scale Hurst exponent of sea clutter frequency spectrum performs better in separating target from sea clutter than the mono-Hurst exponent and the multi-scale Hurst exponent in time domain.Additionally,because Fourier transform can promote the signal-to-clutter ratio effectively,the proposed detection method has the potential for detecting weak moving targets within sea clutter.
摘要:Constrained nonnegative matrix factorization was an excellent method for hyperspectral unmixing.The traditional algorithm of this method was based on projected gradient method,and its convergence rate,accuracy and stability needed to be improved.To this end,we considered the excellent minimum volume constraint,and proposed a fast algorithm for hyperspectral unmixing based on constrained nonnegative matrix factorization.First the minimum volume constrained model of the original problem was optimized,then an alternating direction method of multipliers was used to solve the non-convex constrained nonnegative matrix factorization,and at last we modified the iteration steps by singular value decomposition.Experimental results on simulated and real hyperspectral data demonstrate the superiority of the proposed algorithm.
关键词:nonnegative matrix factorization(NMF);alternating direction method of multipliers;linear spectral unmixing;minimum volume constraint
摘要:Terahertz solid state frequency multiplying sources and sensors are developed with GaAs Schottky barrier diodes and hybrid integrated circuit process.Based on physical structure of diode,high efficiency multipliers,and high sensitivity sensors,such as detectors and subharmonic mixers (SHM) are developed with the combination of electromagnetic (EM) full-wave tool and circuit simulation tool.To the 0.15THz detector,highest measured voltage sensitivity is 1600mV/mW,typical sensitivity is 600mV/mW in 0.11~0.17THz,and tangential signal sensitivity (TSS) is superior than -29dBm.To the 0.15THz frequency doubler,highest measured multiply efficiency is 7.5%,and typical efficiency is 6.0% in 0.1474~0.152THz.To the 0.18THz frequency doubler,highest measured multiply efficiency is 14.8%,and typical efficiency is 8.0% in 0.15~0.2THz.To the 0.15THz SHM,lowest measured conversion loss is 10.7dB,and typical conversion loss is 12.5dB in 0.135~0.165THz.To the 0.18THz SHM,lowest measured conversion loss is 5.8dB,and typical conversion loss is 13.5dB and 11.5dB in 0.165~0.2THz and 0.21~0.24THz,respectively.
摘要:Ultrasonic guided waves were widely used in aeronautics,pipeline,vessels and so on,because of its long distance propagation and large range detection property.Due to the long distance propagation along waveguides,the guided waves were not only possess the feature of ultrasonic bulk waves,but also dispersion and multi-modes characters,which limited detection effect of guided waves and increased the difficulty of feature recognition.It has great theoretical and realistic sense to research the dispersion and multi-modes for ultrasonic guided waves. In this paper,the characters of dispersion and multi-modes were analyzed carefully,and a signal processing point of view was considered.The modeling and morphology decomposition of guided waves signals were proposed by the use of morphology component analysis techniques.The validations of these methods were realized through experiments.All of research results can provide key technologies for decreasing the effect of dispersion and multi-modes and analyzing complex for guided waves signals.
摘要:In this paper,in order to get the minimal relative reduction of features set,heuristic genetic algorithm for feature selection in incomplete decision table is proposed.At first,the fitness function of genetic algorithm is presented.Meanwhile,regarding feature significance as heuristic information in feature selection,and relative core feature serves as initial population to optimize chromosome,which can reduce the exploration space,what’s more,the corresponding condition chromosomes are deleted in the crossover and mutation processes,this method can accelerate the convergence.At last,the better effect of the proposed algorithm can be tested by the experiment.
摘要:In allusion to multi-parameter estimation of a spin echo train (SET) signal in the presence of additive white Gaussian noise,a two-dimensional (2-D) parameter estimation method in a separable manner is proposed.By utilizing the rank-one property of the 2-D data matrix,the damping factor and frequency are estimated in a separable manner from the principal left and right singular vectors according to an iterative weighted least squares (WLS) method.The complex amplitude of SET is then obtained on the basis of standard least squares.The parameter estimation performance of this method achieves Cramer-Rao lower bound (CRLB) at relatively large signal-to-noise ratio (SNR) and/or data size conditions.Simulation results show the effectiveness of the proposed algorithm.
摘要:A novel approach of moving target localization based on the focused imaging is proposed for the problem of "azimuth position uncertainty" in the single-channel SAR (Synthetic Aperture Radar).According to the deduced relationship between moving target and azimuth antenna pattern,the approach uses the relative endpoints of moving target Doppler spectrum to estimate moving target parameters,which can locate the moving target precisely.This approach with lower computation load can be used in lower signal to clutter ratio,which is easy to realize and has wide applications.Both simulation and real data experiments illustrate the effectiveness of this approach.
摘要:How to obtain target speed and location in the case of unknown equivalent electric model and unknown model number of layers of the environment is a difficult problem of the current water weapon guidance technology.Through the analysis of the basic characteristics of the target electrostatic field,a way to obtain the motional target speed and direction in case of the unknown target equivalent electric model is proposed.Through the analysis of the electrical field distribution produced by mixture electric dipole of unknown equivalent model,the electric field normal value of the target symmetrical plane was found to be zero.Therefore,we can achieve model unknown target guidance by finding this plane.1:100 ship model was used in laboratory experiment,and the results of the experiment shows that the target speed and direction of the unknown model can be obtained in this way.The accuracy reached more than 97%,even if the terrain error was introduced,which can meet the requirements of engineering applications.
摘要:The routing performance issues in untrusted environments are the serious challenges for the privacy protection routing protocols in mobile ad-hoc networks (MANET).In order to encourage the selfish nodes and improve the routing performance in untrusted environments,this paper proposes an anonymous incentive mechanism based on hash-chain for the location-aided routing protocols.By using the efficiency and irreversibleness of hash-chain to the location-aided routing protocols,this mechanism can realize immediate incentive to the anonymous forwarders.Moreover,this mechanism can optimize the routing discover process through a route selection mechanism based on the payment price.The anonymity analyses indicate this mechanism can guarantee the anonymities of the nodes involved in routing.The performance evaluation shows that this mechanism has little effect on the routing performance when large amounts of data are transferred,and it is more efficient in small scale networks.
摘要:In higher order quadrature amplitude modulation (QAM) orthogonal frequency-division multiplexing (OFDM) system,the distribution of high-density constellation becomes more sensitive to carrier frequency offset (CFO) and sampling frequency offset (SFO).In this paper,the impact of frequency synchronization errors on high order QAM modulated OFDM signal is analyzed under both time-invariant multipath channel and multipath Rician fading channel,and an algorithm to jointly correct both CFO and SFO is proposed.A closed-form formula for the mean square error (MSE) of the estimation is given by the theoretical analysis and its reasonability is verified by simulation results.
摘要:The existing methods could not identify compound concept effectively from large-scale domain corpus.This paper proposes a method of compound concept extraction based on a hybrid model.Firstly,we make segmentation processing for corpus texts and add entry label for each term.We secondly remove noise words and merge synonyms for the entry set.Then we count the weighted term frequency,the location affinity degree,the location matching degree,and make a stepwise estimation to identify composite concept with atomic terms.Ultimately we realize the extraction of multiple-compound concept via giving different compound depth.On the foundation of the extraction method,we carried out the experiments with different corpora for compound concept extraction.The results indicated the method has higher recall and precision.
关键词:corpus;domain concept;compound concept;weighted term frequency;entry label;location affinity;compound depth
摘要:Baseline length is a significant factor in direction finding.The longer the baseline is,the higher the accuracy is.When baseline length is more than half-wavelength,phase difference between two arrays will be ambiguity.Most existing unwrapping phase ambiguity algorithms limit baseline length.Combined with time of arrival (TOA) estimation,a kind of long baseline unwrapping phase ambiguity algorithm is presented to realize direction finding of narrowband signal,which reduces the restrictive conditions of baseline length.The simulation results show that the probability of unwrapping phase ambiguity is insensitive to baseline length and incident angle.High accuracy can be achieved in direction finding by setting baseline length under the appropriate signal to noise ratio (SNR).
关键词:direction finding;interferometers;unwrapping phase ambiguity;time of arrival (TOA)
摘要:An intuitive systemic model based on the systemic Yoyos and stochastic differential geometry is provided for finding a meaningful geometric description of radar specific emitter identification (SEI) in this paper.We show that there is a lower dimensional state manifold which generates signals with intrinsic signatures in every emitter.Geometric significances of the manifold go far towards solving SEI problems.A recently popularized manifold learning technique,called Diffusion Maps,is said to preserve the local proximity among sampling data points by first representing the underlying manifold.So this paper examines SEI using the technique to extract diffusion features of signal instantaneous parameters for experiments on actual intercepted radar signals with several same type emitters.Finally,results illuminate the validity of features and correctness of the proposed model.
摘要:In this paper the order-range applied by fractional calculus R-L definition,G-L definition and Caputo definition along with the connections between above three definitions are discussed.The differences of fractional-order derivatives and integer-order derivatives are pointed out.An uniform formula of fractional-order integrals and derivatives along with a physical interpretation of fractional calculus are given.The definition of fractional calculus with constant extent of integral and its direct numerical value arithmetic are put forward,and its application is anticipated.
关键词:R-L definition of fractional calculus;G-L definition of fractional calculus;caputo definition of fractional calculus;physical interpretation of fractional calculus;fractional calculus with constant extent of integral
摘要:A data model namely the approximate repeat vector (ARV) is introduced to describe the similar fragments in DNA sequences.By employing bioinformatics features in data preprocessing,and using ARVs in compression codebook's construction,we propose an asymmetric DNA sequence compression algorithm of biological Lempel-Ziv-Markov chain algorithm 2(BioLZMA-2).The particle swarm optimization (PSO) based memetic algorithm improvement namely the comprehensive learning intelligent particle swarm optimization memetic algorithm (CLIPSO-MA) is employed in the compression codebook's design.Experimental results on benchmark sequences demonstrate better performance of BioLZMA-2 than the original DNA sequence compression algorithms.
关键词:DNA sequence data compression;bioinformatics;approximate repeat vector;memetic algorithm
摘要:Large and ultra large scale phased arrays have been more and more widely applied because of its high gain and directiveness in the desired directions.However,the enormous feed networks and receiving/transmitter channels,required by digital beam-forming for the whole array,need huge cost and have poor real time.A novel algorithm,optimized for both subarray division and synthesis subarray weighting of sum and difference patterns,is proposed in the paper,in order to avoid grating lobes and acquire lower sidelobe,deep null of difference patterns and fewer cost of beam gain.The simulation results and performance analysis are given to demonstrate the effectiveness and feasibility of the presented algorithm.
关键词:large and ultra-large scale phased array;subarray division;intelligent optimization algorithm;joint optimization of sum and difference multi beams
摘要:The color feature is the crucial basis for segmenting a stratum image,but traditional segmentation algorithms based on color feature clustering cannot get desirable results because of colors mixture near the layers' boundaries and heavy noisy.A new image segmentation method is proposed,it adopts an interactive image segmentation algorithm based on Random Walks to improve the computing method of the membership functions of fuzzy C-mean(FCM) incorporating spatial information.It inserts labeled cluster-nodes into the graph formed with pixels and their 4-connectedness,and takes the probabilities that a random walker starting its walk at a pixel first reaches a labeled cluster-node as the membership it belongs to this cluster.The experimental results show that pixels in the area of mixed colors near the layers' boundary are classified more accurately;the method decreases the noise sensitivity of FCM,and can effectively segment the structure physical modeling images.
摘要:In the frequency modulated continuous wave(FMCW)squint synthetic aperture radar(SAR)imaging,not only the phases but also the envelopes of the echoes will be affected by the third term in range.If it isn’t well-corrected,the performance of focusing for the SAR image will be deteriorated.Thus,the echo expression in the range-Doppler domain is presented by the series Reversion algorithm when the range expression comes to the third item.After that,a correction is implemented in the range-Doppler domain.In addition,considering the azimuth-variance in phase brought by the linear rang migration correction in azimuth time domain,the non-linear chirp scaling algorithm is utilized to solve the azimuth-variance problem.And a flowchart of the proposed FMCW imaging algorithm with motion compensation for processing the real measured data is given in this paper.Finally,the simulated experiment with multiple point targets scenario and the imaging result of real measured data are utilized to validate the feasibility and effectiveness of the proposed imaging algorithm.
摘要:The waveguide is decomposed into several substructures in the longitudinal direction,and PML absorbing boundary is treated as one of them.The transverse section of the substructures is discretized with the finite element method,a precise integration method based on the Riccati equations is used for the longitudinal integration to generate the export stiffness matrices,and then the waveguide discontinuity problems are expected to be solved through the combination of the substructures.Numerical examples demonstrate that,the waveguide discontinuity problems can be effectively solved with this method,the disadvantage that the artificial boundary must be far away from the non-uniform region can be overcome,and the field near the non-uniform region can be worked out.
关键词:waveguide discontinuity;finite element method;perfectly matched layer;substructure;precise integration
摘要:In order to improve the precision and robustness of parameter estimation,the multi-aspect SAR data is applied to estimate model parameter of attributed scattering center.We consider the parameter estimation as a sparse signal reconstruction problem,and propose parameter-sequential algorithm to relieve the computational complexity.Two factors are studied.First,the aspect and frequency diversity can improve the performance of dictionary matrix.Second,in order to reduce the algorithm complexity,the parameter estimation procedure is realized sequentially.The initial imagery is reconstructed by dictionary matrix which is built up by the ideal point scattering model.The model order,location and type of scattering center are established primarily by energy segment of initial imagery.Then all parameters are estimated over again based on the dictionary matrix which is built up by the prior estimation.The feasibility and robustness of algorithm is validated by numeric simulation.
摘要:In the existing patch sparsity based image inpainting algorithms,the exemplar-size and neighborhood-consistence weight are fixed,and the best match patches of the patch to be filled are searched in the whole source region.However,it decreases the connectivity of structure and clearness of texture while increases the time complexity of this algorithm.To address these problems,an adaptive image inpainting algorithm is proposed based on patch structure sparsity,in the light of the relationship between the characteristics of damage region and patch structure sparsity.According to patch structure sparsity value of the point which has the maximal prority value,the size of patch to be filled,the neighborhood consistence weight and the part-search region size are adaptively confirmed through setting some parameters,then these parameters are analysed and discussed by some experiments.Experimental results show that the PSNR is increased by 0.3~1.2dB and the speed is improved by 3~7 times compared with the existing algorithms.
关键词:image inpainting;patch structure sparsity;sparse representation;consistence of neighborhood
摘要:In order to simplify the selection of sliding coefficient of conventional sliding mode control switch converter,the relations among the sliding region,the sliding coefficient and the circuit parameters of Buck converter are analyzed and discussed.when inductor current is zero,according to the relationship between the systems trajectory and sliding region,an optimal design method for the sliding coefficient was proposed to ensure that under load variation,the switch converter enter the sliding region without the output voltage oscillation phenomena.Based on load variation scope,the optimal sliding coefficient is calculated directly.The simulation and experimental results verify the validity and feasibility of the optimal design method proposed.
摘要:With the prosperity of stereoscopic industry,the disparity-based stereoscopic technology becomes more and more popular.But this technology may bring viewing discomfort and various image distortions such as depth plane curvature,depth non-linearity and so on.To analyze the origins,characteristics and relations of these perceptual issues,we investigated the stereoscopic distortion model and the comfortable viewing zone.This paper first summarized and compared three different configurations of the disparity-based stereoscopic video capturing systems,and focused on the parallel configuration system.The geometry of stereoscopic camera and display systems was presented.The mapping relationship,known as the distortion model,between the camera space and the viewing space was discussed in the parallel-shifting stereoscopic video systems.This distortion model is the basis of the stereoscopic information processing.The shape distortion factor and the depth factor were employed to explain the puppet theater effect and the cardboard effect.The comfortable viewing zone was considered to reduce the problem of visual discomfort and visual fatigue.Its three different representations were compared both qualitatively and quantitatively.At last,the distortion model and the comfortable viewing zone were combined to draw a conclusion of the stereoscopic shooting rules.
摘要:The Chinese Spectral Radioheliograph (CSRH) is a solar-dedicated radio interferometric array that will produce high spatial resolution,high time resolution and high frequency resolution images of the sun simultaneously in cm-dm wave range.The quality of the calibration is the key for imaging especially for the case of ultra-wideband,high time resolution and frequency resolution in design and implementation.In this paper,we choose the sun as calibration source and calculate the delay for different geographical latitude and longitude.As a result,the delay accuracy that we got is better than 1ns.
摘要:The noise sound sources localization method used in shallow water is proposed.The proposed method establishes the array signal model according to the characteristics of the underwater acoustic multi-path channel,and makes use of the multi-path information and the coherent matched idea to construct the spatial focused steering vector which is in accordance with the actual sound propagation characteristics.As a result,the proposed method can not only overcome the multi-path effect but also improve the localization accuracy.Furthermore,we take the ideas of the focused beamforming based on the worst-case performance optimization for reference,and impose constraint conditions on the actual focused steering vectors to improve the optimum weight vectors and spatial spectrum form.It turns out that this method involves a quadratic minimum problem subject to non-linear constraints.Through the discussion and analysis of the spatial spectrum structure,localization errors,-3dB beamwidth,peak and sidelobe ratio (PSR),we present the relationships between the constraint parameter choices and the mismatch errors.It is proved that the proposed algorithm can improve the robustness of the high resolution method MVDR(Minimum Variance Distortionless Response),and obtain sharper spectrum peaks and better noise interference suppression ability.
摘要:A simple and practical method for the calibration of digital array’s receiving channels is proposed.Theoretical analysis demonstrates that the LFM signal of zero intermediate frequency synchronization among the different receiving channels of an antenna depends on their phase and time calibration.The phase calibration of a wideband LFM signal in receiving channels can be converted into the calibration of a single frequency’s.The method of its amplitude calibration and the required measurment precision of the relative delays for receiving channels calibration are also derived.A wideband beamforming procedure is developed by using pulse compression based on channels calibration.The experimental results prove that the method is effective.
摘要:Different from traditional IP switches,the iSCSI virtualization switch not only realizes the IP packets transmission,but also performs the processing of TCP,iSCSI and SCSI protocol.The short packet first scheduling algorithm insures the priority transmission of iSCSI command and status PDUs (Packet Data Unit),but does not impose restrictions on the forwarding bandwidth of such packets.Moreover,it is lack of quantitative analysis on iSCSI data flow characteristics for this method.The mathematical model of iSCSI PDUs forwarding process within the iSCSI virtualization switch was created by queuing theory.Then the packet scheduling algorithm called FC-WFQ (Flow Control-WFQ) was designed for the iSCSI virtualization switch used in the SANs.Finally the iSCSI virtualization switch with FC-WFQ was tested under certain network scenarios built by ns-2.In the experiments,while the command arrival rate and the ratio of read/write commands were changing,the FC-WFQ dynamically adjusted the forwarding bandwidth weight of each packet flow correspondingly.The results show that the scheduling algorithm can greatly reduce the mean response time of the read/write tasks,and significantly improve the throughput of the iSCSI virtualization switch.
关键词:SAN;SCSI;iSCSI virtualization switch;scheduling algorithm;WFQ;queuing model
摘要:Memristor,a nonlinear resistor with memory property,is the fourth basic two-terminal circuit element realized physically recently.In this paper,an active flux-controlled memristor with piecewise-quadratic nonlinearity is taken as an example,upon which voltage-current relationship of the active memristor is analyzed,and frequency response characteristic of a series circuit consisting of an active flux-controlled memristor and a capacitor (simplified by active WC circuit)is studied.The comparison analysis between active WC circuit and active RC circuit is performed.The results indicate that voltage-current characteristics of the active memristor exhibit pinched hysteresis loops and depend on the inner initial states;both active WC circuit and active RC circuit show high pass properties,but the former is a lead network and the latter is a lag network.Circuit simulations are performed based on equivalent circuit of active flux-controlled memristor,which well verify theoretical analysis results.
摘要:From the viewpoint of optimal image contrast measure,a kind of optimization mathematical models and its enhanced methods are proposed for image enhancement by histogram equalization.For the shortcoming of low operation efficiency of optimal contrast image enhancement method by linear programming,the gray level mapping optimization mathematical model for classical histogram equalization is established to improve its mathematical theory.The improved gray level mapping optimization models for histogram equalization are constructed by means of weighted geometry averaging in order to avoid the shortcoming of classical histogram equalization.The good properties of its optimal solution for gray level mapping optimization models are discussed,and the classical and adjustable histogram equalization methods are regarded as special cases of the proposed optimization mathematical models for image enhancement.Experimental results show that the optimization model of histogram equalization is reasonable and can obtain satisfactory image enhancement effect.To some extent,it is more universal than classical and adjustable histogram equalization methods.
摘要:Particle swarm optimization (PSO) hasbeen successfully applied in the design andoptimization of electromagnetic structures.Combined with finite-difference time-domain (FDTD) technique,PSO algorithm was applied in the design of frequencyselective surfaces.Through theuse of graphics processing unit (GPU) acceleration,the FDTD program achieves about 100 times speedup while preserving the broad-band nature of the algorithm.Standard PSO and binary PSO algorithms were implemented to deal with known cell structure and unknown cell structure optimization respectively.The feasibility and efficiency of the proposed optimization scheme areverified by numerical results.
摘要:For the application of multi-sensor multi-target tracking,a method of data association based on ant colony algorithm and genetic algorithm is proposed in this study.First,this method definites pheromone differently for each independent ant entity.Then,improved global pheromone increment model,and combined crossover and variation operation with fitness model of population.The experimental results of actual data demonstrate the presented algorithm is effective.
摘要:In the rocket propulsion stage,only a few Doppler radars can track the target due to Doppler radar networks generally cover a very large area.We could not get the accurate trajectory estimate with insufficient velocity measurements.In this paper,we propose a trajectory estimation approach based on measurements fusion.The proposed method gets a rough trajectory estimate using Doppler radar angle-measurements.The initial trajectory estimate obtained by bearing measurement intersection is taken as the pseudo-measurement.Then,the pseudo-measurements are fused with velocity measurements via unscented Kalman filter (UKF).The Monte Carlo simulation results show that the proposed approach can compensate the accuracy of bearing-only measurement,and improve the track estimation.
摘要:In order to improve the performance of speaker recognition system in noisy environment,this paper presents an auditory feature extraction algorithm.It used adaptive compression Gammachirp filter banks to simulate the auditory characteristics of human cochlea,and the input speech signal was sub-band filtered in the frequency-domain.After logarithmic transformation,it can get the logarithmic sub-band energy as the auditory feature parameter.It respectively used discrete cosine transform and kernel principal component analysis method to transform the auditory feature and get the two new auditory features,which not only can reduce the dimension of the feature parameters,but also can improve the robustness and personality expression of feature parameters.The experimental results show that speaker recognition system with the new auditory feature parameters can get the better results in the robustness and recognition performance than Mel cepstral coefficients and auditory feature parameters based on Gammatone filter banks.
关键词:speaker recognition;adaptive compression Gammachirp filter;kernel principal component analysis;feature extraction