Email alert
1992 Vol. 14, No. 5
column
Display Method:
1992, 14(5): 449-456.
Abstract:
A neural network to implement the maximum likelihood bearing estimation algorithm in real time is proposed. Both analysis and simulation show that this neural network is guaranteed to be stable and to provide the maximum likelihood bearing estimation within an elapsed time of only a few characteristic time constants of the network. As a result, this proposed neural network is satisfactory for real time bearing estimation.
A neural network to implement the maximum likelihood bearing estimation algorithm in real time is proposed. Both analysis and simulation show that this neural network is guaranteed to be stable and to provide the maximum likelihood bearing estimation within an elapsed time of only a few characteristic time constants of the network. As a result, this proposed neural network is satisfactory for real time bearing estimation.
1992, 14(5): 457-464.
Abstract:
The theory of uniform filter banks using all-pass filters is further developed. A filter-bank structure using two-stage all-pass filters is proposed. The pre-stage is half-band filters with period, the post-stage is two pass-band filter banks. The stopband of the pre-ttage filter just controls the don t-care band of the poststage filter using all-pass polyphase, so that a realization of continuous stopband property is obtained Moreover, a method of implementing synthesis filter banks is given, which eliminates aliasing and amplitude distortions of analysis/synthesis system. Finally, an example is listed.
The theory of uniform filter banks using all-pass filters is further developed. A filter-bank structure using two-stage all-pass filters is proposed. The pre-stage is half-band filters with period, the post-stage is two pass-band filter banks. The stopband of the pre-ttage filter just controls the don t-care band of the poststage filter using all-pass polyphase, so that a realization of continuous stopband property is obtained Moreover, a method of implementing synthesis filter banks is given, which eliminates aliasing and amplitude distortions of analysis/synthesis system. Finally, an example is listed.
1992, 14(5): 465-471.
Abstract:
Based on studying deeply the Monolithic Crystal Filter (MCF) theory and itsequivalent circuit, the design and analysis of MCF by computer is described. In the design program, the quasi-empirical equation is given and some coefficients are modified for avoiding to compute transcendental equation. The building block technology and step connecting matrix method are used in analyzing program, so the program procedure becomes more simple and the computation speed is increased. Finally some kinds of monolithic crystal filters are designed and manufaetured by using this program, and good results are obtained.
Based on studying deeply the Monolithic Crystal Filter (MCF) theory and itsequivalent circuit, the design and analysis of MCF by computer is described. In the design program, the quasi-empirical equation is given and some coefficients are modified for avoiding to compute transcendental equation. The building block technology and step connecting matrix method are used in analyzing program, so the program procedure becomes more simple and the computation speed is increased. Finally some kinds of monolithic crystal filters are designed and manufaetured by using this program, and good results are obtained.
1992, 14(5): 472-478.
Abstract:
Several kinds of stream ciphers-complement sequences of period sequences,partial sum of period sequences, inverse order sequences and finitely generated sequences, are studied by using techniques of generating functions. Their minimal polynomials, periods, as well as generating functions are given. As to finitely generated sequences, the change of their linear complexity profiles as well as the relationship between the two generated sequences under the case in which the degree of connected polynomials are fixed, are dicussed.
Several kinds of stream ciphers-complement sequences of period sequences,partial sum of period sequences, inverse order sequences and finitely generated sequences, are studied by using techniques of generating functions. Their minimal polynomials, periods, as well as generating functions are given. As to finitely generated sequences, the change of their linear complexity profiles as well as the relationship between the two generated sequences under the case in which the degree of connected polynomials are fixed, are dicussed.
1992, 14(5): 479-485.
Abstract:
A model of random panicles constructed by the operation of self-similarity in fractal geometry is presented, and the correlation function of its number density is obtained. The model, which can be constructed with a definite method, has the characteristic of self-similarity in structure, and its correlation functions can be obtained accordingly. Therefore, it can be used conveniently in theoretical study and digital simulations of wave interaction in random media. As an example, this model has been applied to analyse the range dependence of volume scattering in radar echoes. The results agree well with Rastogi's (1990) simulation results.
A model of random panicles constructed by the operation of self-similarity in fractal geometry is presented, and the correlation function of its number density is obtained. The model, which can be constructed with a definite method, has the characteristic of self-similarity in structure, and its correlation functions can be obtained accordingly. Therefore, it can be used conveniently in theoretical study and digital simulations of wave interaction in random media. As an example, this model has been applied to analyse the range dependence of volume scattering in radar echoes. The results agree well with Rastogi's (1990) simulation results.
1992, 14(5): 486-495.
Abstract:
A UTD (uniform geometrical theory of diffraction for electromagnetic waves) solution for the field excited by electric and (or) magnetic dipoles on a perfectly conducting cylinder is derived by DGF (dyadic Green s function) technique at first. Then, with the fundamental principle of high-frequency electromagnetic field, the character of geometrical optics and the differential geometry theory, a UTD solution for the radiation field of dipole on a perfectly conducting smooth convex surface is obtained by directly extending the results of the canonical problem. The formulae given in this paper agree primarily with that obtained by P. H. Pathak et al. (1981) except for some factors of dyadic transfer function in lit region. Some factors derived by P. H. Pathak are corrected.
A UTD (uniform geometrical theory of diffraction for electromagnetic waves) solution for the field excited by electric and (or) magnetic dipoles on a perfectly conducting cylinder is derived by DGF (dyadic Green s function) technique at first. Then, with the fundamental principle of high-frequency electromagnetic field, the character of geometrical optics and the differential geometry theory, a UTD solution for the radiation field of dipole on a perfectly conducting smooth convex surface is obtained by directly extending the results of the canonical problem. The formulae given in this paper agree primarily with that obtained by P. H. Pathak et al. (1981) except for some factors of dyadic transfer function in lit region. Some factors derived by P. H. Pathak are corrected.
1992, 14(5): 496-501.
Abstract:
The study of the back scarering from a planar or cylindrical array of loaded dipoles is presented. The current distribution on the dipoles and the radar cross-section (RCS) of the array in consideration of the interaction among dipoles are obtained by the moment method. Theoretical results are compared with the measured ones, and good agreement is achieved. The loading array technique and the curved surface array technique would greatly reduce the RCS of an array, which would be valuable for engineering applications.
The study of the back scarering from a planar or cylindrical array of loaded dipoles is presented. The current distribution on the dipoles and the radar cross-section (RCS) of the array in consideration of the interaction among dipoles are obtained by the moment method. Theoretical results are compared with the measured ones, and good agreement is achieved. The loading array technique and the curved surface array technique would greatly reduce the RCS of an array, which would be valuable for engineering applications.
1992, 14(5): 502-508.
Abstract:
A characteristic analysis of the finlines with uniaxially anisotropic substrates using the spectral-domain immittance approach is presented. The optical axis of uniaxial substrate is either perpendicular or parallel to the substrate plane. This approach is tested to be correct, through the numerical calculation of the effective dielectric constant of a bilateral symmetry finlines.
A characteristic analysis of the finlines with uniaxially anisotropic substrates using the spectral-domain immittance approach is presented. The optical axis of uniaxial substrate is either perpendicular or parallel to the substrate plane. This approach is tested to be correct, through the numerical calculation of the effective dielectric constant of a bilateral symmetry finlines.
1992, 14(5): 509-516.
Abstract:
The Lunar waveguides and vained circular waveguides are analyzed by using the mode matching method. The fields in the waveguides are expressed by cylindrical harmonic expansions in which Bessel functions with half orders are involved and the boundary conditions are imposed. Characteristic equations for cutoff wave numbers for both TE and TM modes are obtained. Several waveguides of different relative dimension are calculated, and the results show high accuracy and are considerably steady.
The Lunar waveguides and vained circular waveguides are analyzed by using the mode matching method. The fields in the waveguides are expressed by cylindrical harmonic expansions in which Bessel functions with half orders are involved and the boundary conditions are imposed. Characteristic equations for cutoff wave numbers for both TE and TM modes are obtained. Several waveguides of different relative dimension are calculated, and the results show high accuracy and are considerably steady.
1992, 14(5): 517-522.
Abstract:
Pt-GaAs Schottky barrier APDs have been investigated. The devices were fabricated on GaAs epitaxial layer with carrier concentration of 0.5-31015cm-3 and thickness of about 20 m. Guard ring along with sensitive area was formed by H+ bombardment with energy of 500 keV and dosage of 11015cm-2 to prevent edge breakdown. Semi-transparem Pt film was evaporated using a special evaporation source. The peak response wavelength of the device is 8600 -8835 at different bias voltages. Optical absopdon edge could extended to 9700 . Franz-Keldysh effect has been observed. The multiplication of above 100 could reach. Dark current is about several nA. Excess noise coefficient is 7 and both rise and fall time were less than 1 ns. The device could be integrated monolithically and planarly with GaAs FET.
Pt-GaAs Schottky barrier APDs have been investigated. The devices were fabricated on GaAs epitaxial layer with carrier concentration of 0.5-31015cm-3 and thickness of about 20 m. Guard ring along with sensitive area was formed by H+ bombardment with energy of 500 keV and dosage of 11015cm-2 to prevent edge breakdown. Semi-transparem Pt film was evaporated using a special evaporation source. The peak response wavelength of the device is 8600 -8835 at different bias voltages. Optical absopdon edge could extended to 9700 . Franz-Keldysh effect has been observed. The multiplication of above 100 could reach. Dark current is about several nA. Excess noise coefficient is 7 and both rise and fall time were less than 1 ns. The device could be integrated monolithically and planarly with GaAs FET.
1992, 14(5): 526-531.
Abstract:
A novel broadband modem wlh high performances suitable for MAP (manufacturing automation protocal) network is introduced It uses the AM/PSK technique. The design isnew and considers not only the related international standards but also the user s requirements.The modem can complete all the physical layer performances of he broadband MAP network and has been used for single-cable broadband MAP networking suvcessfully.
A novel broadband modem wlh high performances suitable for MAP (manufacturing automation protocal) network is introduced It uses the AM/PSK technique. The design isnew and considers not only the related international standards but also the user s requirements.The modem can complete all the physical layer performances of he broadband MAP network and has been used for single-cable broadband MAP networking suvcessfully.
1992, 14(5): 532-536.
Abstract:
The motion compensation for ISAR imaging using scattering centroid (called reference centroid method) is proposed and studied. The internal relations and differences between correlation method and the new method are analyzed. The noise effect on the motion compensation is discussed. To solve the problem, a new method which uses echo goodness-of-fit curve to smooth the centroid range trace and phase trace is proposed. The reference centroid method is used to process the simulative and real ISAR data of movi...
The motion compensation for ISAR imaging using scattering centroid (called reference centroid method) is proposed and studied. The internal relations and differences between correlation method and the new method are analyzed. The noise effect on the motion compensation is discussed. To solve the problem, a new method which uses echo goodness-of-fit curve to smooth the centroid range trace and phase trace is proposed. The reference centroid method is used to process the simulative and real ISAR data of movi...
1992, 14(5): 537-540.
Abstract:
By using an image of snow sample, the correlation function and correlation length were obtained experimentally. Then the effective radius of discrete spherical scat-terer in the radiative transfer model was determined. By solving the vector radiative transfer equation for a slab of densely-distributed scatterers, the polarized brightness temperatures, including multiple scattering effect, from a layer of snow were numerically calculated, and favorably compared with the measurements from spaceborne remote sensing.
By using an image of snow sample, the correlation function and correlation length were obtained experimentally. Then the effective radius of discrete spherical scat-terer in the radiative transfer model was determined. By solving the vector radiative transfer equation for a slab of densely-distributed scatterers, the polarized brightness temperatures, including multiple scattering effect, from a layer of snow were numerically calculated, and favorably compared with the measurements from spaceborne remote sensing.
1992, 14(5): 541-545.
Abstract:
The multistatic frequency swept imaging of point targets based on the authors extended Bojarski s identity is simulated with computer and is compared with that based on S. R. Raz s idea. It is demonstrated that the former is evidently better than the later.
The multistatic frequency swept imaging of point targets based on the authors extended Bojarski s identity is simulated with computer and is compared with that based on S. R. Raz s idea. It is demonstrated that the former is evidently better than the later.
1992, 14(5): 546-549.
Abstract:
On the basis of the spectral domain immittance approach and wavematrix technique, an efficient numerical method for analyzing multi-layer microstrip antennas is presented. The formulas are simple and easy to be programmed. Numerical results are in good agreeement with experimental results, including those of other authors.
On the basis of the spectral domain immittance approach and wavematrix technique, an efficient numerical method for analyzing multi-layer microstrip antennas is presented. The formulas are simple and easy to be programmed. Numerical results are in good agreeement with experimental results, including those of other authors.
1992, 14(5): 550-554.
Abstract:
Two dimensional numerical simulation and analysis for the static state characteristics of i-GaAlAs/i-GaAs HIGFETs by using finite-element method are presented. Some improvements have been made on the boundary conditions, mesh generation and estimation of initial values in the program. The electron concentration and potential distribution etc. inside the HIGFETs are computed. The results of its output characteristics are in good agreement with the experimental data.
Two dimensional numerical simulation and analysis for the static state characteristics of i-GaAlAs/i-GaAs HIGFETs by using finite-element method are presented. Some improvements have been made on the boundary conditions, mesh generation and estimation of initial values in the program. The electron concentration and potential distribution etc. inside the HIGFETs are computed. The results of its output characteristics are in good agreement with the experimental data.
1992, 14(5): 555-560.
Abstract:
For various catalysts at different temperature, the percentages of O2 concentration versus time or number of discharge pulses were obtained in a closed-cycle TEA CO2 laser with the magnetic oxygen analyzer. The experimental phenomena were analysed. Several catalysts were compared in detail.
For various catalysts at different temperature, the percentages of O2 concentration versus time or number of discharge pulses were obtained in a closed-cycle TEA CO2 laser with the magnetic oxygen analyzer. The experimental phenomena were analysed. Several catalysts were compared in detail.
1992, 14(5): 523-525.
Abstract:
The difference between the reciprocity theorem and the mutual energy theorem is discussed, and it is shown that not all linear lossless networks in zero state are reciprocal.
The difference between the reciprocity theorem and the mutual energy theorem is discussed, and it is shown that not all linear lossless networks in zero state are reciprocal.
