Email alert
2000 Vol. 22, No. 3
Display Method:
2000, 22(3): 353-358.
Abstract:
In this paper, the time-frequency characteristics of complex targets are studied and the stability of time-frequency distribution of complex target is obtained with the aspect throughout the large angle range. Based on these table signatures, recognition algorithm is proposed. The experiments of recognition using the data of three kinds of aircraft target models are performed, and demonstrate that there are high recognition performance and robust recognition algorithm.
In this paper, the time-frequency characteristics of complex targets are studied and the stability of time-frequency distribution of complex target is obtained with the aspect throughout the large angle range. Based on these table signatures, recognition algorithm is proposed. The experiments of recognition using the data of three kinds of aircraft target models are performed, and demonstrate that there are high recognition performance and robust recognition algorithm.
2000, 22(3): 359-365.
Abstract:
In this paper an automatic multiresolution classification method is proposed to classify polarimetric synthetic aperture radar (SAR) image.At first a textured SAR image characterizing the terrain reflection is obtained by using a multi-look polarimetric whitening filter (MPWF) to reduce the speckle in the polarimetric SAR image. Then the wavelet transform (WT) is utilized to extract texture information in different resolutions, and in the lowest resolution level the Akaik information criterion (AIC) is used for estimating the optimal number of texture classes in the image. Next the Markov random field (MRF) model is employed to characterize the spatial constraints between pixels in each resolution level, and a maximum like- lihood (ML) approach and an iterated conditional mode (ICM) approach are used for the model parameters estimation and maximum a posteriori (MAP) classification, respectively. Finally the paper presents the experimental results with the NASA/JPL L-band airborne polarimetric SAR data and verifies the effectiveness and advantage of the classification method proposed.
In this paper an automatic multiresolution classification method is proposed to classify polarimetric synthetic aperture radar (SAR) image.At first a textured SAR image characterizing the terrain reflection is obtained by using a multi-look polarimetric whitening filter (MPWF) to reduce the speckle in the polarimetric SAR image. Then the wavelet transform (WT) is utilized to extract texture information in different resolutions, and in the lowest resolution level the Akaik information criterion (AIC) is used for estimating the optimal number of texture classes in the image. Next the Markov random field (MRF) model is employed to characterize the spatial constraints between pixels in each resolution level, and a maximum like- lihood (ML) approach and an iterated conditional mode (ICM) approach are used for the model parameters estimation and maximum a posteriori (MAP) classification, respectively. Finally the paper presents the experimental results with the NASA/JPL L-band airborne polarimetric SAR data and verifies the effectiveness and advantage of the classification method proposed.
2000, 22(3): 366-372.
Abstract:
With the help of optimal performance of the nonlinear least square method,two optimal location algorithms(the LMF location algorithm and quasi-Newton location algorithm) are presented in this paper.Comparting to foregone location algorithms,the simulation results show that the two new location algorithms have better location performance and avoid location ambiguous question without adding new measurements.
With the help of optimal performance of the nonlinear least square method,two optimal location algorithms(the LMF location algorithm and quasi-Newton location algorithm) are presented in this paper.Comparting to foregone location algorithms,the simulation results show that the two new location algorithms have better location performance and avoid location ambiguous question without adding new measurements.
2000, 22(3): 373-378.
Abstract:
In this paper, an algorithm of generating INSAR unwrapped phase image from SAR single-look complex images is presented. Besides the general processing technique, this article focuses on the methods of flat-earth phase removal, phase noise reduction and phase unwrapping. The availability is tested by the results of processing ERS-1/2 SAR images.
In this paper, an algorithm of generating INSAR unwrapped phase image from SAR single-look complex images is presented. Besides the general processing technique, this article focuses on the methods of flat-earth phase removal, phase noise reduction and phase unwrapping. The availability is tested by the results of processing ERS-1/2 SAR images.
2000, 22(3): 379-384.
Abstract:
Various factors leading to error in spaceborne SAR radiometric calibration are discussed,and an equation to estimate the radiometric calibration error is proposed．The equation for calculating SAR imaging processor gain based on chirp scaling algorithm is derived,and testified through computer simulation experiments.A quantitative conclusion on the contribution of imaging processor gain to SAR radiometric calibration error is obtained.
Various factors leading to error in spaceborne SAR radiometric calibration are discussed,and an equation to estimate the radiometric calibration error is proposed．The equation for calculating SAR imaging processor gain based on chirp scaling algorithm is derived,and testified through computer simulation experiments.A quantitative conclusion on the contribution of imaging processor gain to SAR radiometric calibration error is obtained.
2000, 22(3): 385-391.
Abstract:
Monopulse radar three-dimensional(3-D)images coincide with targets'size,but ISAR imaging does not hold,which is importantin precision guidance.The movement character of ship targets and the Doppler frequency induced by targets'3-D rotation are analysed,and dechirping technique to separate scatterers in coherent processing interval is proposed.For angle glint induced in angle measurement by monopulse radar,a method for processing such case is achieved.Simulation results show the high quality of monopulse 3-D images.
Monopulse radar three-dimensional(3-D)images coincide with targets'size,but ISAR imaging does not hold,which is importantin precision guidance.The movement character of ship targets and the Doppler frequency induced by targets'3-D rotation are analysed,and dechirping technique to separate scatterers in coherent processing interval is proposed.For angle glint induced in angle measurement by monopulse radar,a method for processing such case is achieved.Simulation results show the high quality of monopulse 3-D images.
2000, 22(3): 392-397.
Abstract:
The Superresolution mehtod based on the models of both the signal and the noise leads tO a superior accuracy thah the FFT method．This paper presents applications of superresolution signal estimatiors to Doppler beam sharpened imaging and proposes a method for target feature extraction．Experimental results show that the imaging quality are superiorhan that of traditional FFT method.
The Superresolution mehtod based on the models of both the signal and the noise leads tO a superior accuracy thah the FFT method．This paper presents applications of superresolution signal estimatiors to Doppler beam sharpened imaging and proposes a method for target feature extraction．Experimental results show that the imaging quality are superiorhan that of traditional FFT method.
2000, 22(3): 398-401.
Abstract:
A sampling theorem for signals in arbitrary passband is proposed and the efficient algorithm to solve the least sampling rate is given,which can be applied to signal reprdsentation based on its channel pattern.
A sampling theorem for signals in arbitrary passband is proposed and the efficient algorithm to solve the least sampling rate is given,which can be applied to signal reprdsentation based on its channel pattern.
2000, 22(3): 402-410.
Abstract:
This paper first introduces the idea about orientation wavelet subtree quantization and constructs a hybrid image coding using orientation wavelet subtree fractal quantization and zerotree wavelet quantization. With this coding method the horizontal, vertical and diagonal scaling factors of quadtrees are treated independently according to the statistic analysis of test image orientation property and a result of a marked improvement in accuracy with respect to quantization for image compression is obtained. The coding results of this hybrid coding scheme outperform conventional fractal coders and Davis SQS method.
This paper first introduces the idea about orientation wavelet subtree quantization and constructs a hybrid image coding using orientation wavelet subtree fractal quantization and zerotree wavelet quantization. With this coding method the horizontal, vertical and diagonal scaling factors of quadtrees are treated independently according to the statistic analysis of test image orientation property and a result of a marked improvement in accuracy with respect to quantization for image compression is obtained. The coding results of this hybrid coding scheme outperform conventional fractal coders and Davis SQS method.
2000, 22(3): 411-415.
Abstract:
Noise occurring in image generation and communication affects image quality.As a rule,median filtering or weighting smoothing is used to dwindle noise.But those ways can keep limited image details,some more value information also dwindles.This paper proposes a difference-filtering way to solve this problem.It is very important in remote sensing and medicine.
Noise occurring in image generation and communication affects image quality.As a rule,median filtering or weighting smoothing is used to dwindle noise.But those ways can keep limited image details,some more value information also dwindles.This paper proposes a difference-filtering way to solve this problem.It is very important in remote sensing and medicine.
2000, 22(3): 416-422.
Abstract:
For the operator M which is a mean shifted version of IFS(Iterated Function System) in wavelet domain is proved that in practical circumstances the same fixed point can be reached in only few iterations. So, the IFS in wavelet domain may converge without being fully contractive. Then, a contraction factor s upper bound is obtained after relaxing the collage theorem bound and the idea of eventual contractivity introduced by Fisher is explained.
For the operator M which is a mean shifted version of IFS(Iterated Function System) in wavelet domain is proved that in practical circumstances the same fixed point can be reached in only few iterations. So, the IFS in wavelet domain may converge without being fully contractive. Then, a contraction factor s upper bound is obtained after relaxing the collage theorem bound and the idea of eventual contractivity introduced by Fisher is explained.
2000, 22(3): 423-428.
Abstract:
In this papcr,a new approach to blind equalization based on cellular neural network (CNN)is Proposed.According to a reasonable template design method the template for blind equalization is designed.The result of computer simulation proves this approach is reasonable.
In this papcr,a new approach to blind equalization based on cellular neural network (CNN)is Proposed.According to a reasonable template design method the template for blind equalization is designed.The result of computer simulation proves this approach is reasonable.
2000, 22(3): 429-433.
Abstract:
This paper presents a kind of neural network model for nonlinear programming problems on the closed convex sets, defines its Liapunov energy function and shows its global asymptotic stability by using La Salle invariant principle under proper assumptions. Finally, the simulations are given to illustrate the correctness of proposed model.
This paper presents a kind of neural network model for nonlinear programming problems on the closed convex sets, defines its Liapunov energy function and shows its global asymptotic stability by using La Salle invariant principle under proper assumptions. Finally, the simulations are given to illustrate the correctness of proposed model.
2000, 22(3): 434-438.
Abstract:
This paper studies the problem of global exponential stability for the cellular neural networks. Three criterions about global exponential stability of the cellular neural networks are obtained by means of Lyapunov function approach and the method of inequality analysis.
This paper studies the problem of global exponential stability for the cellular neural networks. Three criterions about global exponential stability of the cellular neural networks are obtained by means of Lyapunov function approach and the method of inequality analysis.
2000, 22(3): 439-443.
Abstract:
The binary primitive M-sequences are discussed in this paper. It is shown that arbitrary two reciprocal primitive M-sequences have the same auto-correlation function and the equal linear complexity, meanwhile, the configuration of the two M-sequences is proposed.
The binary primitive M-sequences are discussed in this paper. It is shown that arbitrary two reciprocal primitive M-sequences have the same auto-correlation function and the equal linear complexity, meanwhile, the configuration of the two M-sequences is proposed.
2000, 22(3): 444-450.
Abstract:
A frequency limited orthogonal(FLO)wavelet with expected frequency property is constituted on the basis of Multiresolution Analysis (MRA)．A parameteris defined to characterize the frequency distribution of QRS complexes．Testing on varieties of waveforms in MIT arrhythmia database shows that ventricular ORS complexes usually yield smallerthan that of normal QRS complexes．With such parameter,especially evaluated by the derived FLO wavelet，the stronger capability of classifying ventricular and normal beats can be oberved even in the noise contaminated ECG．
A frequency limited orthogonal(FLO)wavelet with expected frequency property is constituted on the basis of Multiresolution Analysis (MRA)．A parameteris defined to characterize the frequency distribution of QRS complexes．Testing on varieties of waveforms in MIT arrhythmia database shows that ventricular ORS complexes usually yield smallerthan that of normal QRS complexes．With such parameter,especially evaluated by the derived FLO wavelet，the stronger capability of classifying ventricular and normal beats can be oberved even in the noise contaminated ECG．
2000, 22(3): 451-458.
Abstract:
The problem of optimal coherent detection of frequency-shift keying (FSK) signals transmitted over an additive colored Gaussian noise channel with imbalance gains in different frequencies(IGDF-ACGN channel) is studied. The structure of the optimal coherent detector is derived. Theoretical analysis and computer simulation results show that the detector is optimal in either additive colored Gaussian noise environment or additive white Gaussian noise environment.
The problem of optimal coherent detection of frequency-shift keying (FSK) signals transmitted over an additive colored Gaussian noise channel with imbalance gains in different frequencies(IGDF-ACGN channel) is studied. The structure of the optimal coherent detector is derived. Theoretical analysis and computer simulation results show that the detector is optimal in either additive colored Gaussian noise environment or additive white Gaussian noise environment.
2000, 22(3): 459-463.
Abstract:
A scheme of conference connection supported by Clos networks is presented.With the assumption in this scheme,some couples of input-output switch modules of conference network are exclusively assigned to every conference,it is proved that conference non-blocking in wide sense and conference independent routing require least middle switch modules of Clos netwotk I．E．M-3．Furthermore．It is pointed out that this schene has characteristic of conference member non-blocking and conference independency,with advantages of flexibility,stability and low latency．
A scheme of conference connection supported by Clos networks is presented.With the assumption in this scheme,some couples of input-output switch modules of conference network are exclusively assigned to every conference,it is proved that conference non-blocking in wide sense and conference independent routing require least middle switch modules of Clos netwotk I．E．M-3．Furthermore．It is pointed out that this schene has characteristic of conference member non-blocking and conference independency,with advantages of flexibility,stability and low latency．
2000, 22(3): 464-470.
Abstract:
According to the mechanism of Fiber Distributed Data Interface (FDDI), the characters of the access method are analyzed in order to set up a queue model for FDDI. Under lighter traffic, the mathematical analysis of the model is given,and the results of emulation experiments are illustrated.
According to the mechanism of Fiber Distributed Data Interface (FDDI), the characters of the access method are analyzed in order to set up a queue model for FDDI. Under lighter traffic, the mathematical analysis of the model is given,and the results of emulation experiments are illustrated.
2000, 22(3): 471-477.
Abstract:
When three dimensional vector scattering is analyzed by integral equation and the moment method(MM)or the fast multipole method(FMM)，the singular integrand will be done，If it is calculated by Gaussian integration directly,the accuracy will be very poor.In order to get an accurate result，this paper provides a method of integral domain division,which divides the integral domain into several parts with one of them containing the singularity.In the singular part,an analytic result can be derived by simplifying the integral function,replacing the variables,of adding and subtracing a sinlar term．The other can be calculated by Gaussion integration directly.Numerical results for RCS of a corner-reflector antenna and a conducting sphere calculated by this method are in reasonable agereement with that in the references．
When three dimensional vector scattering is analyzed by integral equation and the moment method(MM)or the fast multipole method(FMM)，the singular integrand will be done，If it is calculated by Gaussian integration directly,the accuracy will be very poor.In order to get an accurate result，this paper provides a method of integral domain division,which divides the integral domain into several parts with one of them containing the singularity.In the singular part,an analytic result can be derived by simplifying the integral function,replacing the variables,of adding and subtracing a sinlar term．The other can be calculated by Gaussion integration directly.Numerical results for RCS of a corner-reflector antenna and a conducting sphere calculated by this method are in reasonable agereement with that in the references．
2000, 22(3): 478-484.
Abstract:
The movement of antenna s near-field is studied in time domain. The near field distributions are analyzed by using plane wave-spectrum method. A sampling principle is deduced for planar near-field scanning in time domain. Some methods of near-far field transfer are introduced, which can be performed through time or frequency approach. The characteristics of the two approaches are discussed. Extensive numerical results of an H-plane waveguide horn illustrate the analysis.
The movement of antenna s near-field is studied in time domain. The near field distributions are analyzed by using plane wave-spectrum method. A sampling principle is deduced for planar near-field scanning in time domain. Some methods of near-far field transfer are introduced, which can be performed through time or frequency approach. The characteristics of the two approaches are discussed. Extensive numerical results of an H-plane waveguide horn illustrate the analysis.
2000, 22(3): 485-491.
Abstract:
Research Work on electron optics system of multi-beam klystron has been described systematically in this paper.Design consideration and design methods for periodic reversal permanent magnet focusing system are presented ,special technical problems met in design of multi-beam electron gun with controlling electrode are discussed.2D computer software or magnetic field and electron trajectory of electron optics system has been established,and applied to calculate the electron optics system of S-band 200 kW multi-beam klystron.This multi-beam klystron has output power of 180kW,bandwidth of 7.6％,efficiency of 44.4％,gain of ggreater than 45dB,DC beam transmission of 89%.RF beam ransmission of 69～83%.There is good consistence between experiment and calculation.
Research Work on electron optics system of multi-beam klystron has been described systematically in this paper.Design consideration and design methods for periodic reversal permanent magnet focusing system are presented ,special technical problems met in design of multi-beam electron gun with controlling electrode are discussed.2D computer software or magnetic field and electron trajectory of electron optics system has been established,and applied to calculate the electron optics system of S-band 200 kW multi-beam klystron.This multi-beam klystron has output power of 180kW,bandwidth of 7.6％,efficiency of 44.4％,gain of ggreater than 45dB,DC beam transmission of 89%.RF beam ransmission of 69～83%.There is good consistence between experiment and calculation.
2000, 22(3): 492-495.
Abstract:
The double-barrier light emission tunnel junctions have been fabricated. The characteristics of electronic resonant tunneling with the junctions have been analyzed. In connection with the electronic resonant tunneling and light emission mechanism of the junction, the negative resistance phenomenon (NRP) in the I-V curve and the relation between NRP and Surface Plasmon Polariton(SPP) have been studied especially.
The double-barrier light emission tunnel junctions have been fabricated. The characteristics of electronic resonant tunneling with the junctions have been analyzed. In connection with the electronic resonant tunneling and light emission mechanism of the junction, the negative resistance phenomenon (NRP) in the I-V curve and the relation between NRP and Surface Plasmon Polariton(SPP) have been studied especially.
2000, 22(3): 505-508.
Abstract:
Based on BAN s basic constructs and rules, the freshness in BAN logic is investigated. In this paper it is presented that some reasons why BAN logic is in practice unsound are not idealization problems, but problems what some basic rules result in. For the latter, a modified method is proposed to avoid similar problems.
Based on BAN s basic constructs and rules, the freshness in BAN logic is investigated. In this paper it is presented that some reasons why BAN logic is in practice unsound are not idealization problems, but problems what some basic rules result in. For the latter, a modified method is proposed to avoid similar problems.
2000, 22(3): 509-512.
Abstract:
How to distribute a secret key information to n authorized departments(denoted as: S1, S2Sn, respectively).For every autherized department Si(#em/em#=1, 2,, n), there are qi authorized persons,and every authorized person has the same secret key information as that of any other authorized persons in the same authoried department.The secret key information can be reconstrueted Completely from the secret key information of any k authorized persons， if there is at least one authorized person included for every authorized department.But other cases,the secret key information cannot be reconstructed completely.In this paper,based on the algebraic semigroup theory,a secret key sharing scheme is proposed,which can satisfy the demand.
How to distribute a secret key information to n authorized departments(denoted as: S1, S2Sn, respectively).For every autherized department Si(#em/em#=1, 2,, n), there are qi authorized persons,and every authorized person has the same secret key information as that of any other authorized persons in the same authoried department.The secret key information can be reconstrueted Completely from the secret key information of any k authorized persons， if there is at least one authorized person included for every authorized department.But other cases,the secret key information cannot be reconstructed completely.In this paper,based on the algebraic semigroup theory,a secret key sharing scheme is proposed,which can satisfy the demand.
2000, 22(3): 513-516.
Abstract:
In order to recover the statistically imdependent source signals，the paper proposes a novel blind beam-forming approach based on two-order statistics．In this method，the source signals are extracted by the statistic properties of the zero cross-correlation between any two source signlas，under the case of unknown array manifold．The approach can be applied to the array beamforming with arbitrary array manifold and error．The method requires that the noise is white and the source signals are colored or deternuninistic，and not requires the estimation of arrival-direction of signals．In extracting source signals，it will make SNR be largest．On thecontrary to the blind beamforming besed on higher-order statistics，it is simple and robust，and can extensively be applied to the blind separation of mobile communication signals．
In order to recover the statistically imdependent source signals，the paper proposes a novel blind beam-forming approach based on two-order statistics．In this method，the source signals are extracted by the statistic properties of the zero cross-correlation between any two source signlas，under the case of unknown array manifold．The approach can be applied to the array beamforming with arbitrary array manifold and error．The method requires that the noise is white and the source signals are colored or deternuninistic，and not requires the estimation of arrival-direction of signals．In extracting source signals，it will make SNR be largest．On thecontrary to the blind beamforming besed on higher-order statistics，it is simple and robust，and can extensively be applied to the blind separation of mobile communication signals．
2000, 22(3): 517-521.
Abstract:
The multi-layer perceptron is introduced to characterize the microstrip line．The size of microstrip line and the frequency are defined as the input saples of the multi-layer perceptron，and the electric characteristic parameters of the microstrip，which are obtained by the numerical method，are determined as the desired output samples．As the perceptron has been trained with the back propagation algorithm，the electric characteristic parameters of the microstrip line at any arbitrary size and frequencies can be obtained quickly from the multi-layer perceptron．
The multi-layer perceptron is introduced to characterize the microstrip line．The size of microstrip line and the frequency are defined as the input saples of the multi-layer perceptron，and the electric characteristic parameters of the microstrip，which are obtained by the numerical method，are determined as the desired output samples．As the perceptron has been trained with the back propagation algorithm，the electric characteristic parameters of the microstrip line at any arbitrary size and frequencies can be obtained quickly from the multi-layer perceptron．
2000, 22(3): 522-525.
Abstract:
The solution of the electromagnetic field is given for TE mode in a infinite long fan-shaped waveguide. The basic principle of a rotating mode filter with fan-shaped waveguide is presented. The design method for the filter is discussed by a practical example. The good experimental result of the filter is reached for TE03 mode operation.
The solution of the electromagnetic field is given for TE mode in a infinite long fan-shaped waveguide. The basic principle of a rotating mode filter with fan-shaped waveguide is presented. The design method for the filter is discussed by a practical example. The good experimental result of the filter is reached for TE03 mode operation.
2000, 22(3): 526-528.
Abstract:
In this paper，the impedance matching problem of the circular waveguide latching ferrite variable polarizer has been studied，a method of calculating the impedance matching of the circular waveguide latching ferrite variable polarizer has been presented．Formulas for calculating the /4 impedance matching are given．According to the size obtained by the calculation，a device for impedance matching was made and the VSWR of the device has been measured．The measured results agree well with the calculated results．
In this paper，the impedance matching problem of the circular waveguide latching ferrite variable polarizer has been studied，a method of calculating the impedance matching of the circular waveguide latching ferrite variable polarizer has been presented．Formulas for calculating the /4 impedance matching are given．According to the size obtained by the calculation，a device for impedance matching was made and the VSWR of the device has been measured．The measured results agree well with the calculated results．
2000, 22(3): 496-504.
Abstract:
In recent years,the researches on combination of wavelet analysis and neural net works have attracted much attention.This paper reviews the development and status about this field.The combination of wavelet and neural networks can be categorizde into the ancillary type and the embedded type,the latter of which is referred to wavelet neural networks(WNN). WNN including its main models,algorithms and other issues are discussed.The applications and prospects of WNN are also given,which showed that WNN have great competence and po tential in the applications of function approximation,signal classification,system identification and image compression．
In recent years,the researches on combination of wavelet analysis and neural net works have attracted much attention.This paper reviews the development and status about this field.The combination of wavelet and neural networks can be categorizde into the ancillary type and the embedded type,the latter of which is referred to wavelet neural networks(WNN). WNN including its main models,algorithms and other issues are discussed.The applications and prospects of WNN are also given,which showed that WNN have great competence and po tential in the applications of function approximation,signal classification,system identification and image compression．
