Email alert
1996 Vol. 18, No. 1
Display Method:
1996, 18(1): 1-6.
Abstract:
The attraction domains of memory patterns and the exponential convergence rate of the network trajectories to memory patterns for continuous feedback associative memory are estimated. These results can be used for evaluation of error-correction capability and the synthesis procedures for continuous-time associative memory neural networks.
The attraction domains of memory patterns and the exponential convergence rate of the network trajectories to memory patterns for continuous feedback associative memory are estimated. These results can be used for evaluation of error-correction capability and the synthesis procedures for continuous-time associative memory neural networks.
1996, 18(1): 7-14.
Abstract:
The piecewise linear saturation characteristics of cell in a cellular neural network(CNN) and phase plane analysis method are used to realize linear separable and nonseparable Boolean expressions. And the principle is also used to achieve some CNN smoothing algorithms for binary images.
The piecewise linear saturation characteristics of cell in a cellular neural network(CNN) and phase plane analysis method are used to realize linear separable and nonseparable Boolean expressions. And the principle is also used to achieve some CNN smoothing algorithms for binary images.
1996, 18(1): 15-19.
Abstract:
A new method of deconvolution in the frequency domain is proposed. The relationship between the proposed method and traditional Discrete Fourier Transform(DFT) method is also discussed. It can be seen from this paper that the new method based on the DFT techniques can overcomes the difficulties of using traditional frequency methods to do deconvolution, and the deconvolution problem can always be solved with the new method because three are only DFT, IDFT and multiplication calculations. An example is also presented to illustrated the calculation of the method.
A new method of deconvolution in the frequency domain is proposed. The relationship between the proposed method and traditional Discrete Fourier Transform(DFT) method is also discussed. It can be seen from this paper that the new method based on the DFT techniques can overcomes the difficulties of using traditional frequency methods to do deconvolution, and the deconvolution problem can always be solved with the new method because three are only DFT, IDFT and multiplication calculations. An example is also presented to illustrated the calculation of the method.
1996, 18(1): 20-24.
Abstract:
This paper proposes a new block matching criterion called the bit-correlation matching function for sequence image coding. By using the same fast search algorithm, the bit-correlation matching function not only results in as nearly same a uracy in displacement estimation as mean square error function, but also makes the fast search algorithm low in computation complexity and easy to parallel operation, thus reducing the coding time of sequence image efficiently.
This paper proposes a new block matching criterion called the bit-correlation matching function for sequence image coding. By using the same fast search algorithm, the bit-correlation matching function not only results in as nearly same a uracy in displacement estimation as mean square error function, but also makes the fast search algorithm low in computation complexity and easy to parallel operation, thus reducing the coding time of sequence image efficiently.
1996, 18(1): 25-30.
Abstract:
This paper presents a new motion estimation algorithm for video conference signal coding. This type of algorithm is called block adaptive recursive algorithm (BARA). Simulation results show that this new algorithm has better performance than that of conventional ones.
This paper presents a new motion estimation algorithm for video conference signal coding. This type of algorithm is called block adaptive recursive algorithm (BARA). Simulation results show that this new algorithm has better performance than that of conventional ones.
1996, 18(1): 31-35.
Abstract:
The motion of a relativistic electron is analyzed in the field configuration consisting of a circular wiggler magnetic field, an axial magnetic field, and the equilibrium self-electric and self-magnetic fields produced by the non-neutral electron beam. By generating Poincare surface-of-section maps, it is shown that when the self-fields are strong enough, the electron motions become chaotic. Although the realistic circular wiggler field makes the equations of electron motion nonintegrable as the self-fields do, the effect of self-fields to cause the chaoticity is stronger than the wiggler field. The axial magnetic field can suppress the occurrence of chaoticity.
The motion of a relativistic electron is analyzed in the field configuration consisting of a circular wiggler magnetic field, an axial magnetic field, and the equilibrium self-electric and self-magnetic fields produced by the non-neutral electron beam. By generating Poincare surface-of-section maps, it is shown that when the self-fields are strong enough, the electron motions become chaotic. Although the realistic circular wiggler field makes the equations of electron motion nonintegrable as the self-fields do, the effect of self-fields to cause the chaoticity is stronger than the wiggler field. The axial magnetic field can suppress the occurrence of chaoticity.
1996, 18(1): 36-42.
Abstract:
Based on the analysis of SAR radar return model, an approach using Doppler-rate matching filter is proposed. It can solve the problem of azimuth defocus because of the quadratic phase mismatch. Then a fast algorithm structure given by matching matrix is presented. The computer simulation results illustrate its availability.
Based on the analysis of SAR radar return model, an approach using Doppler-rate matching filter is proposed. It can solve the problem of azimuth defocus because of the quadratic phase mismatch. Then a fast algorithm structure given by matching matrix is presented. The computer simulation results illustrate its availability.
1996, 18(1): 43-49.
Abstract:
The foundational principle of the nonorthogonal PD-TD algorithm is introduced, and its stability condition and dispersion equation are derived. The numerical dispersion characteristics of this method are analyzed in detail. In order to show its validity the scattering surface currents of perfectly conducting cylinder are calculated by using this method when the incident wave is sinusoidal plane wave in TM case. The numerical results agree with that by moment method very well. Then the TM scattering problem of polygon cylinder coated with absorbing material is studied and the monostatic RCS of the target is obtained. For comparison with results of equivalent source method the electric scattering distribution pattern of square cylinder coated with absorbing material is also obtained by using nonorthogonal FD-TD.
The foundational principle of the nonorthogonal PD-TD algorithm is introduced, and its stability condition and dispersion equation are derived. The numerical dispersion characteristics of this method are analyzed in detail. In order to show its validity the scattering surface currents of perfectly conducting cylinder are calculated by using this method when the incident wave is sinusoidal plane wave in TM case. The numerical results agree with that by moment method very well. Then the TM scattering problem of polygon cylinder coated with absorbing material is studied and the monostatic RCS of the target is obtained. For comparison with results of equivalent source method the electric scattering distribution pattern of square cylinder coated with absorbing material is also obtained by using nonorthogonal FD-TD.
1996, 18(1): 50-57.
Abstract:
A finite-difference frequency-domain(FD-FD) method in non-orthogonal coordinate is presented for the eigenvalue problems of waveguide. The numerical discrete mesh of the algorithm can fit the arbitrary boundary well. In this paper the cut-off frequencies of modes of waveguides are calculated via given eigenvalue equation. The results show good agreement with those from other sources.
A finite-difference frequency-domain(FD-FD) method in non-orthogonal coordinate is presented for the eigenvalue problems of waveguide. The numerical discrete mesh of the algorithm can fit the arbitrary boundary well. In this paper the cut-off frequencies of modes of waveguides are calculated via given eigenvalue equation. The results show good agreement with those from other sources.
1996, 18(1): 58-63.
Abstract:
Based on transmission line theory, the voltage transfer ratio function matrixes are derived for the multi-conductor cable system. Then an optimising object function for reducing crosstalk in the multi-conductor cable is established. The nonlinear programming is applied to optimize the terminal impedance load. According to the optimization results, a new method for reducing crosstalk is presented. Theoretical results are in good agreement with experimental ones.
Based on transmission line theory, the voltage transfer ratio function matrixes are derived for the multi-conductor cable system. Then an optimising object function for reducing crosstalk in the multi-conductor cable is established. The nonlinear programming is applied to optimize the terminal impedance load. According to the optimization results, a new method for reducing crosstalk is presented. Theoretical results are in good agreement with experimental ones.
1996, 18(1): 72-76.
Abstract:
The features of non-voice signals on public communication nets are analysed and an approach to distinguish such signals is introduced in this paper. It is the first time for the system to meet three key targets: large-scale, high recognition rate and low false alarm at the same time.
The features of non-voice signals on public communication nets are analysed and an approach to distinguish such signals is introduced in this paper. It is the first time for the system to meet three key targets: large-scale, high recognition rate and low false alarm at the same time.
1996, 18(1): 77-81.
Abstract:
A channel model specially applied to the terrestrial broadcasting of high definition tele-vision(HDTV) is presented on the base of analyzing some kinds of channel models over VHF/UHF. The model sufficiently concerns main interferences faced by HDTV signal in terrestrial broadcasting path. It is very helpful to researches of HDTV transmission system.
A channel model specially applied to the terrestrial broadcasting of high definition tele-vision(HDTV) is presented on the base of analyzing some kinds of channel models over VHF/UHF. The model sufficiently concerns main interferences faced by HDTV signal in terrestrial broadcasting path. It is very helpful to researches of HDTV transmission system.
1996, 18(1): 82-86.
Abstract:
A method of moving adaptive picture interpolation is presented. The interlaced scanning that structures the pictures of today s television systems is changed to progressive scanning, thus vertical resolution of pictures is improved greatly.
A method of moving adaptive picture interpolation is presented. The interlaced scanning that structures the pictures of today s television systems is changed to progressive scanning, thus vertical resolution of pictures is improved greatly.
1996, 18(1): 87-89.
Abstract:
A new millimeter-wave integrated oscillator and design method are introduced in this paper. By combining finline and microstrip transmission line, this oscillator has more than 3.0 GHz mechanical tuning bandwidth and over 70.0mW output power.
A new millimeter-wave integrated oscillator and design method are introduced in this paper. By combining finline and microstrip transmission line, this oscillator has more than 3.0 GHz mechanical tuning bandwidth and over 70.0mW output power.
1996, 18(1): 90-94.
Abstract:
Radar cross-section of targets varies with their configurations, dimensions, wavelength and radiation direction of waves. In this paper, a general representation for the radar cross-section of simply face-shaped scatterer with variance in these parameters is given. With this representation, the radar cross-section of prototype of this scatterer can be evaluated by means of model-testings, even though the scaling relations between the prototype and the models are not satisfied.
Radar cross-section of targets varies with their configurations, dimensions, wavelength and radiation direction of waves. In this paper, a general representation for the radar cross-section of simply face-shaped scatterer with variance in these parameters is given. With this representation, the radar cross-section of prototype of this scatterer can be evaluated by means of model-testings, even though the scaling relations between the prototype and the models are not satisfied.
1996, 18(1): 95-98.
Abstract:
A RCS model testing method for lossy radar targets fully or partially coated by absorbent materials is presented with its basical theory and programing technique. It is shown that this method is usable and practicable because the computed result by model testing agrees well with the measured data.
A RCS model testing method for lossy radar targets fully or partially coated by absorbent materials is presented with its basical theory and programing technique. It is shown that this method is usable and practicable because the computed result by model testing agrees well with the measured data.
1996, 18(1): 99-103.
Abstract:
The Kirchhoff s faltung formula of complex aperture is proposed as a new method for radiation from a plane aperture. It is based on the idea of analytic extension that the original aperture field is extended to a complex plane, and then the Kirchhoff s faltung formula makes up an indirect and rigorous representation of the radiation from a plane aperture. Numerical calculations comfirm its effectiveness and correctness.
The Kirchhoff s faltung formula of complex aperture is proposed as a new method for radiation from a plane aperture. It is based on the idea of analytic extension that the original aperture field is extended to a complex plane, and then the Kirchhoff s faltung formula makes up an indirect and rigorous representation of the radiation from a plane aperture. Numerical calculations comfirm its effectiveness and correctness.
1996, 18(1): 104-108.
Abstract:
Boron ions have been implanted into Si at an incident energy of 3MeV to a dose of 51015cm-2. Buried conductive layers are formed in Si substrate after annealing at 1050℃ for 20s. Annealing characteristics of damage has been examined by double-crystal X-ray diffraction. By detailed theoretical analysis of character of high energy ion implantation and optical response of free-carrier plasma effects, electrical activaton of implanted boron ions has been investigated by computer simulation of the IR reflection interference spectra.
Boron ions have been implanted into Si at an incident energy of 3MeV to a dose of 51015cm-2. Buried conductive layers are formed in Si substrate after annealing at 1050℃ for 20s. Annealing characteristics of damage has been examined by double-crystal X-ray diffraction. By detailed theoretical analysis of character of high energy ion implantation and optical response of free-carrier plasma effects, electrical activaton of implanted boron ions has been investigated by computer simulation of the IR reflection interference spectra.
1996, 18(1): 109-112.
Abstract:
This paper analyzes the mechanism models of metallic impurity infection and adsorption-desorption, and the distribution law of microdefect in growing of silicon CVD epitaxy, and then a new epitaxy technology is proposed, which is optimized by using contrary compensation, and by which the metallic impurity and microdefect in silicon epitaxy layer are reduced effectively.
This paper analyzes the mechanism models of metallic impurity infection and adsorption-desorption, and the distribution law of microdefect in growing of silicon CVD epitaxy, and then a new epitaxy technology is proposed, which is optimized by using contrary compensation, and by which the metallic impurity and microdefect in silicon epitaxy layer are reduced effectively.
1996, 18(1): 64-71.
Abstract:
A new type of high power microwave-multibeam klystron has been described. The development of multibeam klystron is commented on briefly, the principle and main characteristics of multibeam klystron based on the fundamental mode of klystron cavity are given in detail. The effect of the parameters of multibeam klystron on the operating bandwidth and output power has also bean studied, and the power limitation is given in this paper. Finally, the key technical problems and trend of multibeam klystron are pointed out.
A new type of high power microwave-multibeam klystron has been described. The development of multibeam klystron is commented on briefly, the principle and main characteristics of multibeam klystron based on the fundamental mode of klystron cavity are given in detail. The effect of the parameters of multibeam klystron on the operating bandwidth and output power has also bean studied, and the power limitation is given in this paper. Finally, the key technical problems and trend of multibeam klystron are pointed out.
