摘要:In the paper,analyses the time and frequency domain characteristics of the target signal under the classical magnetic dipole model and points out the difficulty and particularity of the magnetic localization and parameter estimation problem.Designs the second-order difference matrix method (SDMM) and frequency domain matching method (FDMM) to solve the localization and parameter estimation problem.Both of the algorithms make use of the time and frequency domain characteristic of the signal and convert a high-order optimization problem into sequential solutions of several low-order problems.SDMM has low computation burden,excellent real-time performance.FDMM is more reliable in low SNR environment and suitable for post-analysis.Simulation of both methods meets well with the theoretical analysis and the algorithms achieve the goal.The paper also points out the further work for the two algorithms.
摘要:The chaotic system will bifurcate when the system parameters change a little.If the measured signal was put to the chaotic system,it would be get the stable state,period states and chaotic states.In the article,controlling the system work at the period states makes the states changes with the signal,A chaotic circuit has been added to the measurement system to improve the delicacy.The measurement delicacy was analyzed.The results of chaotic system immuned with the noises were given.
LI Hong-liang, LIU Gui-zhong, HOU Xing-song, LI Yong-li, ZHANG Zong-ping
Vol. 31, Issue 1, Pages: 8-12(2003)
摘要:Nonlinear multiresolution signal decomposition schemes is a new tool developed over the past years.The lifting scheme introduced by Sweldens also provides a useful way to construct nonlinear wavelet decompositions freely.In this paper,a new method of nonlinear wavelet decomposition is proposed,which is suited for the task of image compression,especially for lossless coding applications.It is based on a certain statistical operator defined here.It can be used to realize the integer-valued wavelet transform,which can avoid quantization with the image detail signals being zero (or almost zero) at the smooth graylevel variation areas at big probability.Numerical results show that the entropy of the coefficients in the transform domain obtained with this new method is smaller than that obtained with other nonlinear transform methods.
摘要:Due to non-orthogonality of signature sequence,performance of linear detector is much less than that of the optimum detector that is nonlinear.Structural risk introduced in SVMs leads to more generalization and less training samples than the other learning models and the nonlinear SVMs can approach optimum multi-user detector.In this paper,SVMs are used for multi-user detection.Considering the variation with time of the wireless channel,an adaptive SVM is presented for multi-user detection.The results of simulation demonstrate the feasibility and validity of this method.
摘要:A novel chaotic communication scheme combining the continuous chaotic system with the digital communication channel is proposed,in which the chaotic modulation and demodulation module are realized with software and the continuous chaotic signals carrying the hidden information are transmitted in digital form.The designing method of the proposed communication scheme is described in detail and the optimal selections of the related parameters to balance the performances of the system in security,transmission rate and transmission quality are discussed.The proposed communication system is easy to practically realize and compared with the conventional digital communication system,the performance of the transmission rate and transmission quality is not lowered while the security of information transmission is obtained.
关键词:chaos;secure communication;digital communication
摘要:This paper puts up a method suitable for multi-pitch detecting under overlapping speech signals environment.In this method,wavelet transform is used as filtering analysis part of this pitch detecting model.Besides that,generalized autocorrelation function is used to strengthen pitch information and enhanced summary autocorrelation function is used to weaken redundant information.It is the most important that a pitch probability function is given to predict and tail after each pitch tracking to improve the veracity of pitch detecting.Above-mentioned method could be applied to computational auditory scene analysis.From the experiment results provided,we can infer that this method is very useful and efficient.
XU Jing, WANG Shu-xun, WANG Hong-zhi, Cheng shi-xin
Vol. 31, Issue 1, Pages: 25-28(2003)
摘要:In this paper,the harmonic retrieval and frequency-coupled harmonics are studied under the condition that the complex noise include multiplicative and additive noise are correlative.First,the concept of cross-mixing is proposed.Second,the result on the frequency-coupled harmonics are extended.Third,the harmonic retrieval in the complex noise is solved with the slice of the sixth-order time-averaged moment polyspectral.The above algorithms are confirmed by the simulations.
ZHU Hui-ling, HANG Da-ming, MA Zheng-xin, CAO Zhi-gang, Victor
Vol. 31, Issue 1, Pages: 29-32(2003)
摘要:QoS Routing is one of the key technologies to provide quality of service in future networks.The goals of QoS Routing are to select feasible paths that meet QoS constraints and to make efficient utilization of the network resource.Metrics selection,feasible paths search and node state information accuracy are three key issues in QoS Routing.On these three issues,this paper presents problems and solutions in QoS Routing,and discusses the possible future directions in the QoS routing area.
关键词:QoS routing;multiple metrics;overhead of paths search;inaccurate state information
摘要:Speaker adaptation is a powerful means of improving the performance of speaker-independent speech recognition system.This paper focuses on state-of-the-art of speaker adaptation technologies that include the modes and algorithms of speaker adaptation.The details of MLLR and MAP algorithms that are widely used are also given.The trend of study on speaker adaptation is predicted in this paper.
摘要:This paper presents an improvement on real-valued genetic algorithm that utilizes a new crossover strategy to steer the location of offspring in genetic algorithm.The use of such a crossover strategy provides the capacity of quickly moving offspring toward regions with improved fitness.Experimental results demonstrate that improved real-valued genetic algorithm could greatly accelerate the speed on obtaining optimal solutions.
XU Bing-ji, LIAO Xiao-xin, LIU Xin-zhi, CHEN Qin-sheng
Vol. 31, Issue 1, Pages: 41-44(2003)
摘要:Global stability of equilibrium point of a class of second order Hopfield neural networks is discussed in this paper.Some sufficient conditions for the global asymptotic stability and global exponential stability of equilibrium point of such networks are obtained by means of Lyapunov functions and linear matrix inequalities(LMI'S),the exponential convergence rate of equilibrium point is also estimated.
关键词:second order neural networks;global asymptotic stability;Exponential convergence rate
摘要:In this paper,idea on basic immunogenetics is introduced to optimization.By simulating the behavior of biological immune system,we design the immunogenetic algorithm.Experiments have been carried out on the function optimization,the results show that the method can improve the optimization capability of Simple Genetic Algorithm(SGA)successfully and so the satisfaction is proved.
摘要:In this paper,the priority control method,called "Hot-Spot Push Out" (HSPO),is used for analyzing the performance of shared buffer ATM switches under a non-uniform traffic.The system performance model is theoretically analyzed and the important calculating formulas of the system performance parameters,such as "cell loss ratio" 、 "mean cell delay" 、 "throughput",etc,are derived based on queueing theory.The system performance is examined by numerical analysis.In addition,The results compared with simulation show that the theoretical analysis is correct and reasonable.
摘要:A new signature scheme,i.e.a vector space secret sharing-multisignature scheme,is proposed based on vector space secret sharing scheme and multisignature scheme.The security of this scheme is analyzed.In this scheme,the group signature can be easily produced if an authorized subset of participants pool their secret shares and it is impossible for them to generate a group signature if an unauthorized subset of participants pool their secret shares.The validity of the partial signature and the group signature can be verified by means of verification equations.A group signature of authorized subset of participants cannot be impersonated by any other set of participants.Moreover the suspected forgery can be traced and the malicious participants can be caught in this scheme.None of several possible attacks can successfully break this scheme.
关键词:vector space secret sharing scheme;multisignature scheme;discrete logarithm;security analysis
摘要:In this paper,we propose an efficient on-line secret sharing scheme using one-way hash function.The proposed scheme provides the property to share multiple secrets and allows participants to be added/deleted dynamically,without having to redistribute new shares.It is able to detect and identify cheaters even if majority of the participants are dishonest,and is more computationally efficient than previous schemes.Therefore,it can be applied to multi-party protocols,group-oriented cryptosystems and etc.
摘要:The aim of image restoration is to achieve super-resolution and increase signal-noise ratio.Poisson-ML image restoration algorithm (PML) has high super-resolution performance,but introduces oscillatory artifacts in the restored image and can not get a ideal restoration result for the noisy image.Super-resolution image restoration algorithm based on Poisson and Markov model as well as the adaptive choice method of the regularization parameter (MPML) is proposed through the assumption of Poisson and Markov random field.Experiments demonstrate that MPML not only has high super-resolution performance,but also can effectively reduce and remove oscillatory artifacts in restored images,and get ideal restoration result for the noisy image.The significantly improved restoration results are obtained using MPML compared with PML.The regularization parameter can be automatically and optimally chosen in step with the restoration of the degraded image.
关键词:image processing;image restoration;super-resolution;Markov random field;Poisson model;Bayes analysis
摘要:According to the geometry deformation characteristic of scanned sequence image used in three-dimensional computer radiotherapy treatment planning system,a sequence image normalizing method based on scanning image rulers has been developed.The method first models the image deformation caused by the tilted film as a projection model,and makes use of CT and MRI sequence image rulers,both vertical and horizontal,to rectify the deformation and calculate a few user-defined marker's relative coordinates; then employs these markers to normalize the whole sequence image.The clinical result shows that the method is feasible and can satisfy the clinical image normalizing demand in three-dimensional radiotherapy treatment planning system.
关键词:image rulers;projection model;sequence image normalizing;three-dimensional radiotherapy treatment planning system
摘要:Digital watermarking is a key technique for protecting intellectual property of digital media.As a number of methods have been proposed in recent years to embed watermarks in images for various applications,evaluation of watermarking algorithms becomes more and more important.Watermark visibility,which can be measured by signal to noise ratio (SNR) or peak signal to noise ratio (PSNR),is one of the major performance indexes of watermarking algorithms.In this paper,based on the Laplacian distribution model of AC DCT coefficients,we deduce theoretical relationship between the scaling parameter in some typical watermarking algorithms and the degradation of watermarked images.Experimental results show that the evaluated error of SNR and PSNR is less than 1 dB.
关键词:digital watermark;evaluation of the visibility;laplacian distribution;signal to noise ratio;peak signal to noise ratio
摘要:In this paper,we propose an algorithm for the semiautomatic segmentation of medical image series based on the combination of the live wire algorithm and the active contour model.We modify the traditional live wire algorithm by combining it with the region growing method and obtain accurate segmentation of one or more slices in a medical slice series by the live wire algorithm.Then the computer will segment the nearby slice using the active contour model.We introduced a gray-scale model to the boundary points of the active contour model to record the local region characters of the desired object in the segmented slice and replace the external energy of the traditional active contour model with the energy decided by the gray-scale model.A simple method based on the idea of the live wire algorithm is introduced for the reparation of the automatic segmentation result to guarantee the reliability of the result.Experiment shows that our algorithm can obtain the boundary of the desired object from a series of medical images quickly and reliably with only little user intervention.
SUN Hong-bo, GUO Xin, GU Hong, SU Wei-min, LIU Guo-sui
Vol. 31, Issue 1, Pages: 78-81(2003)
摘要:The Discrete Chirp-Fourier Transform (DCFT) proposed recently is an effective technique for chirp signal detection.But it requires some rigorous restrictions on the signal length and parameters.These requirements restrict the application of DCFT.In this paper,the definition of DCFT is modified.No restrictions are included in this new definition.It is an excellent and practical technique for the detection of chirp signal.Then the modified DCFT is applied to airborne SAR moving target detection and good results are achieved.In addition,a new CLEAN technique combined with the modified DCFT is proposed to detect the small targets submerged by the large ones effectively.This new combination technique is valuable in practice.The simulation results prove its validity.
摘要:Geodesic active contour model is an effective method for the contour detection of multiple objects on an image.However,there are some limitations and shortcomings in this model,such as poor convergence to boundary concavities,leaking from broken boundary,miserable anti-noise ability,etc.In this paper,based on a detail analysis of geodesic active contour,a new algorithm combined the attraction field regularization and multiscale method is presented to overcome these shortcomings.The regularization of the attraction field extends the attraction range and smoothes the field that can decrease the noise influence.The multiscale method provides an accurate contour detection.The final experiments prove that this new model is sensitive to concave boundary,insensitive to broken boundary,and robust to noise.
关键词:geodesic active contour;contour extraction;multiscale image;object tracking
摘要:The information is carried through the hopping of frequency in DFH(Differential frequency hopping).If only the frequency is detected out,the information can be recovered.One method of hop-detecting based on STFT is put forward,and the frequency features of HF DFH signal in one-hop time is analyzed through it in this paper.Combining with the G function,it can detect the possible frequency.The simulation shows the detecting result is good in several kinds of HF channel.
摘要:This paper introduces fractal theory into the research of multipath fading channels,and a novel Fractal Brownian Motion (FBM) model is established.By studying the increments' probability distribution,statistical self-similarity,generalized power spectra and fractal dimension of multipath fading signals,it shows the inherent relationship between the multipath fading and FBM,and that fractal dimension is a key parameter for describing the signal propagation in wireless channels.Finally,we apply FBM model to the reconstruction of multipath signals,yielding reasonably accurate replication.Simulation results indicate that the FBM model could describe multipath fading phenomena more accurately and effective than the conventional statistical models.
摘要:The introduction of MPLS into Internet makes the Internet traffic engineering possible.It is important to balance the traffic load among several parallel explicit LSPs so as to avoid congestion and optimize the network capacity.The total traffic delay is the most valuable and practicable optimization object for traffic engineering.By analysing the constraint relationship between average packet delay and traffic load,the existence of optimum solution is proved and its mathematical result is gotten.Anovel adaptive traffic contribution method based on such optimization object is proposed in the paper.Both theoretical analyses and simulation bear out the merits of simpleness.speediness and convergence.
摘要:A Multi-camera Human-face Personal Identification method,based on the BIOMIMETIC PATTERN RECOGNITION,had been proposed.The basic theory of neural networks with MULTI-WEIGHT NEURON,in Multi-camera Human-face Personal Identification System Based on the BIOMIMETIC PATTERN RECOGNITION,had been discussed.A three-camera system takes five combined pictures as learning samples in experiments.The basic course of human-face picture preprocessing had been described.The results of experiments showed the correct identification rate as 96%,rejection rate as 4%,and no mistake in identification.Experiments showed that increasing the complexity of neural networks makes better identification result.
摘要:By employing spectral-correlation analysis method,the problem of optimally filtering the cyclostationary signals is discussed in this paper.Based on the max-output-SNR criterion,the analytic expression for the cyclic correlation matched filter (CCMF) is derived,which is just the same as the conventional matched filter except for different signal models.But since the filtering performance for CCMF is primarily related to the selected cycle frequency,which is not unique for most cyclostationary signals,single-cycle CCMF shares the disadvantage of utilizing the incomplete signal information.CCMF bank utilizing multi-cycle frequencies is therefore studied and an optimum structure for the CCMF bank is developed,also based on the max-output-SNR criterion.Simulations on AM signals and BPSK signals are also performed to determine the performance advantage between the spectral-correlation method and cyclic correlation matched filtering method.Simulation results verify the performance for the proposed methods.
关键词:cyclostationary signals;optimal filtering;filter bank
TONG Zhi, TAO Feng, WEI Huai, WANG Mu-guang, WANG Zhi, JIAN Shui-sheng
Vol. 31, Issue 1, Pages: 109-116(2003)
摘要:In this paper we develop a complete model to simulate the properties of a long-period fiber grating whose cladding index is lower than the ambient refractive index based on coupled-mode theory and radiation-mode theory,and we also present a new method to solve the radiation-mode coupling equations numerically.The characteristics of coupling from core-mode to radiation-mode in a LPG is analysed,and the general rules are summarized,which agrees well with experimental results.In addition,this model can be easily applied to fiber Bragg gratings.
关键词:long period fiber grating;radiation mode;leaky-mode;three-layer step-index waveguide
摘要:The hybrid technique of moment method,conjugate gradient method,and fast Fourier transform is applied to analyze 2D electromagnetic scattering from a dielectric cylinder with arbitrary inhomogeneity and anisotropy under oblique-incidence.In the scheme,the integro-differential equation for the total electric field is used.In this way,the size of original question is not enlarged,and the linear system can be efficiently solved by a combination of the conjugate gradient method with the fast Fourier transform.Some numerical results are given in this paper,which demonstrate validity of the scheme.
摘要:An analysis is made on the difficulty of genetic algorithms,in which the convergence is strongly dependent on the selection of initial population.A measure is proposed to alleviate the dependence of convergence on the initial population,in which two reproduction operations of different types are alternatively used in generating chromosoms.The model of improved genetic algorithm for the pattern synthesis of ultra-low side lobe linear array antenna is established.A worked out example is given to demonstrate that the independence of convergence on the selection of initial population is realized in the improved genetic algorithm,showing its good application potentials.
LIU Ji-yuan, TANG Jin-song, SUN Bao-shen, ZHANG Chun-hua, LI Qi-hu
Vol. 31, Issue 1, Pages: 131-134(2003)
摘要:Motion compensation is the key technique for synthetic aperture sonar (SAS),and it is normally realized in SAS systems based on multi receiving arrays and their cross correlation.A new motion compensation method based on single receiving array and strong scatterer is presented in the paper.Moreover,computer simulations based on the derived model for sway error are given,and results of simulation and lake test show the validity of the proposed motion compensation method.
关键词:synthetic aperture sonar;motion compensation;correlation function;displaced phase center
摘要:In this paper,we present a novel HVS-based visual masking and self-adaptive image public watermarking using chaotic array,and define a new noisesensitivity using forth-order cumulant in visual masking.In this algorithm,chaotic array is embedded to the sub-low frequency coefficients of the block-DCT,then visual masking is made according to the noisesensitivity.Under the affect of noisesensitivity,Visual masking accomplish self-adaptive controllable intensity of watermark.Without original image,we can remarkably detect the existence of watermark by public key.With common signal distortion or geometric distortion (such as lossy compress,cropping etc.),watermark can achieve better robust.
摘要:Under motion condition,the channel of Wireless Optical Communication (WOC) suffers from the serious time-varied、space-varied phase.The ability adaptive to the change of WOC channel is so important that we will discuss the blind identification and channel estimate herein.Based on the Second-Order Statistics,denoising,channel estimate of special WOC receiver-transmitter (RT) configuration indoor are obtained in our simulations.The simulated range of SNR is high up to 30dB.The main errors under those WOC condition,are the survivals of singular value decomposition of autocorrelation function and the denoising,which obtained by our detail analysis.The estimate results which coincide with oral channel quantities in different SNR condition is achieved also.
关键词:indoor infrared WOC;blind identification and estimation;time-varied characteristics
摘要:The multipath signal DOA(Direction of Arrival) estimations in mobile communication are discussed in this paper.Some typical algorithms are reviewed in brief first,and the model of multipath signals received by array antennas is described.The algorithm on multipath signal DOA estimation based on time averaging is introduced;also,a problem existing in this method is pointed out;an effective method for solving this problem is raised as well.Finally,the simulation results verifying the theoretical analysis are presented.
关键词:DOA estimation;multipath;array signal processing;mobile communication
DENG Suo-yun, HU Zheng-ming, NIU Xin-xin, YANG Yi-xian
Vol. 31, Issue 1, Pages: 149-153(2003)
摘要:The problem of identity authentication and key agreement is a very important one about security in wireless communications.We propose a wireless mutual authentication and key agreement protocol to resolve this problem.In the protocol,a mobile user can be provided with anonymous service and neither Visited Networks nor wiretappers can get the information of the user's real identity,which guarantees the confidentiality of information of the user's identity and location.This protocol can achieve intended service in an efficient and practical way.
摘要:Although affine model has more ability to describe the motion field,it's very difficult to be used for it's high computational complexity.In this paper,a novel fast algorithm is proposed which is derived from traditional phase-correlation motion estimation method.In FFT frequency domain,the phase shift indicates the translation components of affine motion and the spectrum magnitude indicates the linear components of affine motion.So we utilize this FFT spectrum characteristic to separate the affine parameters estimation to translation components and linear components.The computational complexity is decreased dramatically.For improving the robust of linear component estimation,we use an affine-invariant feature to calculate the correlative points between magnitude spectrums.Simulation indicates that our method can improve the estimation result greatly with sustainable computational complexity.