ISSN 
2006 Vol. 35, No. 3
2006, 35(3): 289-291.
Abstract:
Feature extraction is an important part in handwritten numeral recognition. Efficient and robust feature is a key to improving recognition rate and efficiency. This paper adopts fractional Fourier transform and principal component analysis to extract robust and compact features called fractional eigenafeatures. In classification, five kernel-based nonlinear classifiers, Parzen and robust Parzen classifiers, radial basis function classifier, support vector classifier, and kernel-based nonlinear representor are applied and compared. Experimental results show the effects and efficiency of the proposed algorithm.
Feature extraction is an important part in handwritten numeral recognition. Efficient and robust feature is a key to improving recognition rate and efficiency. This paper adopts fractional Fourier transform and principal component analysis to extract robust and compact features called fractional eigenafeatures. In classification, five kernel-based nonlinear classifiers, Parzen and robust Parzen classifiers, radial basis function classifier, support vector classifier, and kernel-based nonlinear representor are applied and compared. Experimental results show the effects and efficiency of the proposed algorithm.
2006, 35(3): 292-294,316.
Abstract:
To analyse the chaotic time sequences by computer should consider the finite word length of the computer. As to the logistic map, this paper studies the effect of finite word length to its chaotic characters. Due to the finite word length of computer, the chaotic sequences will turn into periodic ones after short interim. The maximum Lyapunov exponents are calculated by the small data sets method. The results show that both periodic and transitional chaotic sequences have positive maximum Lyapunov exponent, and the small data sets method is robust to the finite word length of the computer. In the end, a coupled method to improve the period and interim length of chaotic sequences is given.
To analyse the chaotic time sequences by computer should consider the finite word length of the computer. As to the logistic map, this paper studies the effect of finite word length to its chaotic characters. Due to the finite word length of computer, the chaotic sequences will turn into periodic ones after short interim. The maximum Lyapunov exponents are calculated by the small data sets method. The results show that both periodic and transitional chaotic sequences have positive maximum Lyapunov exponent, and the small data sets method is robust to the finite word length of the computer. In the end, a coupled method to improve the period and interim length of chaotic sequences is given.
2006, 35(3): 295-297,327.
Abstract:
In this paper, a new cost function based on the diagonalization of correlation matrices is proposed to perform blind source separation. This cost function can restrain cross-correlation of separated signals and be applicable to separate instantaneous or convolutive mixture of stationary or non-stationary signals. A real coded genetic algorithm is proposed to search the optimum solution. Computer simulation results demonstrate this algorithm has not only fast convergence performance but also high accuracy.
In this paper, a new cost function based on the diagonalization of correlation matrices is proposed to perform blind source separation. This cost function can restrain cross-correlation of separated signals and be applicable to separate instantaneous or convolutive mixture of stationary or non-stationary signals. A real coded genetic algorithm is proposed to search the optimum solution. Computer simulation results demonstrate this algorithm has not only fast convergence performance but also high accuracy.
2006, 35(3): 298-301.
Abstract:
In this paper, we give a thoroughly study on the approach to design of linear time-invariant filter with minimal output noise enhancement and output signal shape constrained to lie within prescribed upper and lower bounds envelope. With the shape envelope inequality constraints and optimal value conditional expression, the performance requirements of filter can be mathematically formulated as optimization problem. Then the envelope constrained optimal filter design problem can be transform into an QP problem with a convex quadratic cost function and semi-infinite number of linear inequality constraints. Utilizing Lagrangian's dual and Carathé odory's dimensionality theorem, the semi-infinite convex quadratic programming problem is converted into an equivalent finite optimization dual problem, which can be solved by generic optimization methods. Moreover, iterative algorithm for solving equivalent finite optimization dual problem is presented. By a design example, the theory and method are proved most feasible.
In this paper, we give a thoroughly study on the approach to design of linear time-invariant filter with minimal output noise enhancement and output signal shape constrained to lie within prescribed upper and lower bounds envelope. With the shape envelope inequality constraints and optimal value conditional expression, the performance requirements of filter can be mathematically formulated as optimization problem. Then the envelope constrained optimal filter design problem can be transform into an QP problem with a convex quadratic cost function and semi-infinite number of linear inequality constraints. Utilizing Lagrangian's dual and Carathé odory's dimensionality theorem, the semi-infinite convex quadratic programming problem is converted into an equivalent finite optimization dual problem, which can be solved by generic optimization methods. Moreover, iterative algorithm for solving equivalent finite optimization dual problem is presented. By a design example, the theory and method are proved most feasible.
2006, 35(3): 302-304,342.
Abstract:
The OBS length adaptive assembly algorithm can select length thresholds adaptively depending on the input traffic, and make output data bursts length change more slowly than the given length threshold algorithm. But the length thresholds can not be changed by the input self-similar traffic, then it can not depress the self-similarity and packet blocking rate largely. Based on this algorithm, a modified algorithm based on length-changeable thresholds is proposed. The new algorithm can reduce the self-similarity more effectively and adjust the length thresholds according to the properties of the input network traffic stream more adaptively than the length adaptive assembly algorithm. The simulation results show that the packet block rate of the modified adaptive assembly algorithm is only about one tenth of that of adaptive assembly algorithm.
The OBS length adaptive assembly algorithm can select length thresholds adaptively depending on the input traffic, and make output data bursts length change more slowly than the given length threshold algorithm. But the length thresholds can not be changed by the input self-similar traffic, then it can not depress the self-similarity and packet blocking rate largely. Based on this algorithm, a modified algorithm based on length-changeable thresholds is proposed. The new algorithm can reduce the self-similarity more effectively and adjust the length thresholds according to the properties of the input network traffic stream more adaptively than the length adaptive assembly algorithm. The simulation results show that the packet block rate of the modified adaptive assembly algorithm is only about one tenth of that of adaptive assembly algorithm.
2006, 35(3): 305-308.
Abstract:
The frequency sequence detection and error-correcting decoding are analyzed in Differential Frequency Hopping (DFH) communication system. From the viewpoint of the third party of communications, the receiving and decoding process and the complexity of transition function analysis are discussed. The results show that DFH is not a LPI communication manner, and its security relies on the transition function. The complexity of transition function analysis is very high, which provides very high security to DFH communication system.
The frequency sequence detection and error-correcting decoding are analyzed in Differential Frequency Hopping (DFH) communication system. From the viewpoint of the third party of communications, the receiving and decoding process and the complexity of transition function analysis are discussed. The results show that DFH is not a LPI communication manner, and its security relies on the transition function. The complexity of transition function analysis is very high, which provides very high security to DFH communication system.
2006, 35(3): 309-312.
Abstract:
On the basis of the protocol and features of signal-propagating of GSM system, a design project of the MLSE maximum likelihood sequence estimation receiver with soft-output structure of GSM system is proposed. The implementation algorithm of the MLSE receiver is discussed in detail. The test result gained in practical environment shows the MLSE receiver has better performance and potential applications in practical.
On the basis of the protocol and features of signal-propagating of GSM system, a design project of the MLSE maximum likelihood sequence estimation receiver with soft-output structure of GSM system is proposed. The implementation algorithm of the MLSE receiver is discussed in detail. The test result gained in practical environment shows the MLSE receiver has better performance and potential applications in practical.
2006, 35(3): 313-316.
Abstract:
The time-interleaved ADC structure is one way to increase the sampling rate. However, channel mismatch errors of this structure will increasingly degrade the performance of the whole system. For this reason, adding test signal is widely used when measuring and calibrating the mismatch errors. This paper propose a method which measure mismatches without the need of any special calibration signal and therefore has wide applications. Furthermore, the method of error estimation of blind algorithm is improved. The computer simulations show that the method estimates errors with high accuracy and can improve the performance of the system effectively.
The time-interleaved ADC structure is one way to increase the sampling rate. However, channel mismatch errors of this structure will increasingly degrade the performance of the whole system. For this reason, adding test signal is widely used when measuring and calibrating the mismatch errors. This paper propose a method which measure mismatches without the need of any special calibration signal and therefore has wide applications. Furthermore, the method of error estimation of blind algorithm is improved. The computer simulations show that the method estimates errors with high accuracy and can improve the performance of the system effectively.
2006, 35(3): 317-319,330.
Abstract:
This article discusses in detail the character of Inverse Synthetic Aperture Radar (ISAR) multi-target imaging techniques and time-frequency algorithms. Based on the model of rigidity multi-target of multiple ballistic missiles, time-frequency analysis technology is proposed to simulate imaging of multiple targets. Also, Range-Doppler (RD) and three ISAR time-frequency algorithms for multi-targets imaging are compared. The result is given to prove the feasibility of ISAR imaging and recognition of multiple targets.
This article discusses in detail the character of Inverse Synthetic Aperture Radar (ISAR) multi-target imaging techniques and time-frequency algorithms. Based on the model of rigidity multi-target of multiple ballistic missiles, time-frequency analysis technology is proposed to simulate imaging of multiple targets. Also, Range-Doppler (RD) and three ISAR time-frequency algorithms for multi-targets imaging are compared. The result is given to prove the feasibility of ISAR imaging and recognition of multiple targets.
2006, 35(3): 320-323.
Abstract:
In the ground short range target detection of airborne forward-looking arrays, the clutter Doppler frequency highly depends on range. A Doppler compensation method based on vector similarity criteria is proposed, which can estimate the compensation values from the received clutter data. The Doppler compensation, as a pre-processing of space time adaptive processing (STAP) corresponding to clutter data of various range bins, can be utilized for efficient solution of the ground short range clutter nonhomogeneous problem. The method in the pulse domain has the same impact as in Doppler domain and is suitable for parallel implementation; moreover, it has the advantage of low complexity. The simulation shows that a high performance improvement factor can be obtained when it is applied to the pre-processing of STAP.
In the ground short range target detection of airborne forward-looking arrays, the clutter Doppler frequency highly depends on range. A Doppler compensation method based on vector similarity criteria is proposed, which can estimate the compensation values from the received clutter data. The Doppler compensation, as a pre-processing of space time adaptive processing (STAP) corresponding to clutter data of various range bins, can be utilized for efficient solution of the ground short range clutter nonhomogeneous problem. The method in the pulse domain has the same impact as in Doppler domain and is suitable for parallel implementation; moreover, it has the advantage of low complexity. The simulation shows that a high performance improvement factor can be obtained when it is applied to the pre-processing of STAP.
2006, 35(3): 324-327.
Abstract:
In this paper,design and experiment results for a hybrid integrated detector whose center frequency is 35 GHz with 4 GHz bandwidth are presented.The input port of the detector is the form of standard rectangular waveguide.The waveguide to microstrip transition is realized by antipodal finline.By using high frequency electromagnetic simulation and microwave circuit CAD software,the design of the detector is accomplished.
In this paper,design and experiment results for a hybrid integrated detector whose center frequency is 35 GHz with 4 GHz bandwidth are presented.The input port of the detector is the form of standard rectangular waveguide.The waveguide to microstrip transition is realized by antipodal finline.By using high frequency electromagnetic simulation and microwave circuit CAD software,the design of the detector is accomplished.
2006, 35(3): 328-330.
Abstract:
In this paper, a spot-focusing lens antenna operating at 8 mm frequency band is designed according to the measurement of the density of the plasma. The feasibility and validity of the scheme is demonstrated by means of simulation and test of the system. The theory in which high order modes can be ignored is powerfully testified. In addition, some academic and reference value for engineering design of antennas is also presented here.
In this paper, a spot-focusing lens antenna operating at 8 mm frequency band is designed according to the measurement of the density of the plasma. The feasibility and validity of the scheme is demonstrated by means of simulation and test of the system. The theory in which high order modes can be ignored is powerfully testified. In addition, some academic and reference value for engineering design of antennas is also presented here.
2006, 35(3): 331-334.
Abstract:
In this paper, the Uniform geometrical Theory of Diffraction(UTD)is first time employed in conjunction with approximate models of the aircraft structure to analyze the effect of aircraft for double-ridged horn pattern on the top fuselage and to analyze the scattering field formulas effected by aircraft structure. The scattering fields are computed while the radiation pattern of double-ridged horn antenna mounted on the top of an aircraft is obtained. The simulation results show that the effect of the aircraft on radiation pattern is the surface reflecting field, the edge scattering field, and the surface creep wave which contributes to the backload of the pattern. The results of this paper can be used to analyze the radiation pattern of airborne antenna and to optimize the airborne antenna disposition.
In this paper, the Uniform geometrical Theory of Diffraction(UTD)is first time employed in conjunction with approximate models of the aircraft structure to analyze the effect of aircraft for double-ridged horn pattern on the top fuselage and to analyze the scattering field formulas effected by aircraft structure. The scattering fields are computed while the radiation pattern of double-ridged horn antenna mounted on the top of an aircraft is obtained. The simulation results show that the effect of the aircraft on radiation pattern is the surface reflecting field, the edge scattering field, and the surface creep wave which contributes to the backload of the pattern. The results of this paper can be used to analyze the radiation pattern of airborne antenna and to optimize the airborne antenna disposition.
2006, 35(3): 335-338.
Abstract:
In this paper,we first analyze the theory of digital detection and the rationale of CORDIC algorithm. On the basis of this,we propose the design scheme of digital detector based on CORDIC algorithm. The design scheme is successfully implemented by FPGA in the project of VXI all-digital IF real-time wideband RF spectrum analyzer design. The experiment shows that the digital detector based on CORDIC algorithm working with digital down converter has high precision and wide dynamic range:±0.01 dB frequency response flatness and 100 dB dynamic range. If the digital down converter works with high frequency resolution FFT based on DSP, the frequency resolution can reach 0.03 Hz.
In this paper,we first analyze the theory of digital detection and the rationale of CORDIC algorithm. On the basis of this,we propose the design scheme of digital detector based on CORDIC algorithm. The design scheme is successfully implemented by FPGA in the project of VXI all-digital IF real-time wideband RF spectrum analyzer design. The experiment shows that the digital detector based on CORDIC algorithm working with digital down converter has high precision and wide dynamic range:±0.01 dB frequency response flatness and 100 dB dynamic range. If the digital down converter works with high frequency resolution FFT based on DSP, the frequency resolution can reach 0.03 Hz.
2006, 35(3): 339-342.
Abstract:
To simplify the performance analysis of IEEE 802.11DCF in a saturated wireless LAN environment, a simple box-ball model is proposed. It regards the problem of packet transmission collisions as throwing some balls into several boxes randomly and independently. Balls are equal to the number of active nodes and boxes are related to the average contention window of each node. When two or more balls happen to fall into the same box, it means a collision would occur when the corresponding time slot arrives. In concern of limited retransmissions for every packet, the state transform of transmitting node is approximated as a finite-state one-dimension Markovian process. A simple method is finally proposed to evaluate the capacity of 802.11DCF based on the concepts of virtual transmission period and average contention window. Simulation results validate the analysis.
To simplify the performance analysis of IEEE 802.11DCF in a saturated wireless LAN environment, a simple box-ball model is proposed. It regards the problem of packet transmission collisions as throwing some balls into several boxes randomly and independently. Balls are equal to the number of active nodes and boxes are related to the average contention window of each node. When two or more balls happen to fall into the same box, it means a collision would occur when the corresponding time slot arrives. In concern of limited retransmissions for every packet, the state transform of transmitting node is approximated as a finite-state one-dimension Markovian process. A simple method is finally proposed to evaluate the capacity of 802.11DCF based on the concepts of virtual transmission period and average contention window. Simulation results validate the analysis.
2006, 35(3): 343-345.
Abstract:
The performance of the Extensive Bark Spectral Distortion EBSD measure is examined for speech data with coding distortions by varying the scaling the noise masking threshold and omitting the spreading function in noise masking threshold calculation. Some comparative experiment results of the EBSD and Bark Spectral Distortion(BSD) show that the EBSD has better performance than BSD in Chinese communication channels, and the correlation coefficient reaches 0.969.
The performance of the Extensive Bark Spectral Distortion EBSD measure is examined for speech data with coding distortions by varying the scaling the noise masking threshold and omitting the spreading function in noise masking threshold calculation. Some comparative experiment results of the EBSD and Bark Spectral Distortion(BSD) show that the EBSD has better performance than BSD in Chinese communication channels, and the correlation coefficient reaches 0.969.
2006, 35(3): 346-348,374.
Abstract:
This paper gives the software scheme of the front-end under the condition of experiment. The data of the front-end bring UCL (Uniform Content Locator) information automatically, and then is packed as the given format, parallel broadcasted. The design scheme is implemented by four key parts:Uniform Content Locator, parallel broadcast, data packing and controlling of broadcasting. The result of experiment indicates that culture grid is feasible and effective.
This paper gives the software scheme of the front-end under the condition of experiment. The data of the front-end bring UCL (Uniform Content Locator) information automatically, and then is packed as the given format, parallel broadcasted. The design scheme is implemented by four key parts:Uniform Content Locator, parallel broadcast, data packing and controlling of broadcasting. The result of experiment indicates that culture grid is feasible and effective.
2006, 35(3): 349-351.
Abstract:
The global threshold may not work well for image segmentation with small objects. The image segmented by single dynamic threshold usually contains interference like shadow. The gray image segmentation is a process that the image is labeled after an unsupervised clustering by Fuzzy C-Means(FCM). In consideration without increase of computation complexity, we propose an algorithm for gray image segmentation by creating a threshold matrix obtained by global threshold and dynamic threshold that are the FCM clustering results.
The global threshold may not work well for image segmentation with small objects. The image segmented by single dynamic threshold usually contains interference like shadow. The gray image segmentation is a process that the image is labeled after an unsupervised clustering by Fuzzy C-Means(FCM). In consideration without increase of computation complexity, we propose an algorithm for gray image segmentation by creating a threshold matrix obtained by global threshold and dynamic threshold that are the FCM clustering results.
2006, 35(3): 352-355.
Abstract:
In order to overcome the limit of Single-Mode Approximation (SMA), this paper adopts a new method that when expressing the fields inside the grooves, the higher orders are retained as an infinite sum of standing eigen-waves. The dispersion relation is obtained by field-matching method, while the coupling impedance is subsequently obtained from the relevant equations. A lot of numerical computations and analysis are done for rectangular grating slow-wave system. And the effects of the geometrical characteristics of the grating are analyzed. Finally, rectangular waveguide gratings at 3 cm band and 8 mm band are designed and measured, experimental values match theoretical values very well.
In order to overcome the limit of Single-Mode Approximation (SMA), this paper adopts a new method that when expressing the fields inside the grooves, the higher orders are retained as an infinite sum of standing eigen-waves. The dispersion relation is obtained by field-matching method, while the coupling impedance is subsequently obtained from the relevant equations. A lot of numerical computations and analysis are done for rectangular grating slow-wave system. And the effects of the geometrical characteristics of the grating are analyzed. Finally, rectangular waveguide gratings at 3 cm band and 8 mm band are designed and measured, experimental values match theoretical values very well.
2006, 35(3): 356-358.
Abstract:
Based on character of host condensator, degree of super-cooling of vacuum-condensed water is researched. System design rule, model and realization method are illustrated in detail. The system hardware is implemented by Digital Signal Processing (DSP) chip to ensure real-time performance of system, while software is formed by fuzzy-neural networks to overcome uncertainty of system model. Simulation results show that the designed system is stable and reliable with strong robustness.
Based on character of host condensator, degree of super-cooling of vacuum-condensed water is researched. System design rule, model and realization method are illustrated in detail. The system hardware is implemented by Digital Signal Processing (DSP) chip to ensure real-time performance of system, while software is formed by fuzzy-neural networks to overcome uncertainty of system model. Simulation results show that the designed system is stable and reliable with strong robustness.
2006, 35(3): 359-362.
Abstract:
This paper presents the design and implementation of a system that performs remote control of household appliances using personal communication terminals and a household gateway. The system contains a S32410 as the main control module with powerful audio processing capability and a high capacity flash memory which enables telephone to control the household appliances by the gateway and sends message to telephone in case of emergency by gateway. The system's real-time properties and flexibility are increased. Simultaneously, the system can be used more universally and conveniently.
This paper presents the design and implementation of a system that performs remote control of household appliances using personal communication terminals and a household gateway. The system contains a S32410 as the main control module with powerful audio processing capability and a high capacity flash memory which enables telephone to control the household appliances by the gateway and sends message to telephone in case of emergency by gateway. The system's real-time properties and flexibility are increased. Simultaneously, the system can be used more universally and conveniently.
2006, 35(3): 363-366.
Abstract:
Based on the analysis of the characteristics of the test paper, a theoretical model of intelligent test paper generation system is established. Then the method of test paper auto-generation based on Adaptive Hybrid Evolutionary Mechanism (AHEM) is proposed. The experimental results show that new method can solve the problem of test paper auto-generation effectively and efficiently.
Based on the analysis of the characteristics of the test paper, a theoretical model of intelligent test paper generation system is established. Then the method of test paper auto-generation based on Adaptive Hybrid Evolutionary Mechanism (AHEM) is proposed. The experimental results show that new method can solve the problem of test paper auto-generation effectively and efficiently.
2006, 35(3): 367-370.
Abstract:
This paper analyzes the characteristics of distributed Micro Electromechanical System (MEMS) and studies components of MEMS based on multi-agent technology. Then, the control architecture of distributed MEMS designed with multi-agent technology is presented. The system model, structure of micro electromechanical agent, and behavior decision are expressed in formal approach. An algorithm is developed for the controlling the behavior decision and production in distributed real-time MEMS. And a prototype system is realized based on the algorithm.
This paper analyzes the characteristics of distributed Micro Electromechanical System (MEMS) and studies components of MEMS based on multi-agent technology. Then, the control architecture of distributed MEMS designed with multi-agent technology is presented. The system model, structure of micro electromechanical agent, and behavior decision are expressed in formal approach. An algorithm is developed for the controlling the behavior decision and production in distributed real-time MEMS. And a prototype system is realized based on the algorithm.
2006, 35(3): 371-374.
Abstract:
This paper introduces a single-chip-microcomputer control system for supersonic-wave fluid-surface-height measurement and the hardware circuit and the software flow chart of the control system. In order to achieve high precision of measurement,the supersonic-wave sensor is set under the container to make supersonic-wave stable. And the integrated circuit 8254 is used in the system, its CLK is provided by an oscillation circuit of high frequency stability. The measurement result shows that the precision of the system reaches centimeter.
This paper introduces a single-chip-microcomputer control system for supersonic-wave fluid-surface-height measurement and the hardware circuit and the software flow chart of the control system. In order to achieve high precision of measurement,the supersonic-wave sensor is set under the container to make supersonic-wave stable. And the integrated circuit 8254 is used in the system, its CLK is provided by an oscillation circuit of high frequency stability. The measurement result shows that the precision of the system reaches centimeter.
2006, 35(3): 375-377,422.
Abstract:
Cooperation of different industry supervisor systems seamlessly and transparently through Internet is a main trend in industry control systems. Low communication speed, inconsistent protocol and closed system are the primary reasons hindering the development of traditional RTU. Analyzing embedded Internet technology and IEC 60870-5-104 protocol, we propose a general RTU design framework-Webit-GRTU. Application example validates that the design is an effective way for traditional RTU to upgrade.
Cooperation of different industry supervisor systems seamlessly and transparently through Internet is a main trend in industry control systems. Low communication speed, inconsistent protocol and closed system are the primary reasons hindering the development of traditional RTU. Analyzing embedded Internet technology and IEC 60870-5-104 protocol, we propose a general RTU design framework-Webit-GRTU. Application example validates that the design is an effective way for traditional RTU to upgrade.
2006, 35(3): 378-380.
Abstract:
According to the problems that currently many Active Queue Management schemes are heuristic algorithms and lack of systemic and theoretic analysis. In this paper, the optimal control theory is introduced to optimize the Active Queue Management schemes. A mathematical model on the Active Queue Management is given. This model includes two difference equations, which denote the evolvement of queue length and average queue length respectively. Under certain target function, using the drop probability to be the control variable, the problem of optimizing the Active Queue Management can be convert into a dynamic programming problem. Finally, the optimal approach are proposed.
According to the problems that currently many Active Queue Management schemes are heuristic algorithms and lack of systemic and theoretic analysis. In this paper, the optimal control theory is introduced to optimize the Active Queue Management schemes. A mathematical model on the Active Queue Management is given. This model includes two difference equations, which denote the evolvement of queue length and average queue length respectively. Under certain target function, using the drop probability to be the control variable, the problem of optimizing the Active Queue Management can be convert into a dynamic programming problem. Finally, the optimal approach are proposed.
2006, 35(3): 381-384.
Abstract:
Based on the analysis of RTP/RTCP protocol, The RTP data format used for MPEG-4 media is analyzed. An implementation scheme of client using MPEG-4 media data is then proposed. Adopting the scheme that video streams are synchronized to audio stream produces a high playback quality of the media player. The thread model and main architecture of the client and are introduced. Applications show that the clients implemented by the scheme have good stability.
Based on the analysis of RTP/RTCP protocol, The RTP data format used for MPEG-4 media is analyzed. An implementation scheme of client using MPEG-4 media data is then proposed. Adopting the scheme that video streams are synchronized to audio stream produces a high playback quality of the media player. The thread model and main architecture of the client and are introduced. Applications show that the clients implemented by the scheme have good stability.
2006, 35(3): 385-388.
Abstract:
Policy based management is an effective approach to manage distributed system and large scale system. With the basic concepts of policy, domain and policy target introduced, a general policy deployment model is purposed which based on domain layer architecture. It assumes that policies are stored in the most close domain to target, and the whole system changes according to the increment arithmetic. The general model is independent of the underlying concrete policy enforcement mechanisms and can be employed in mixed policy environments.
Policy based management is an effective approach to manage distributed system and large scale system. With the basic concepts of policy, domain and policy target introduced, a general policy deployment model is purposed which based on domain layer architecture. It assumes that policies are stored in the most close domain to target, and the whole system changes according to the increment arithmetic. The general model is independent of the underlying concrete policy enforcement mechanisms and can be employed in mixed policy environments.
2006, 35(3): 389-391,395.
Abstract:
This paper presents a UML-based method for XML Schema design. The method includes three phases:conceptual, logical and physical level design. Standard UML class diagram is utilized in conceptual level, customized logical UML class diagram is utilized in logical level and textual XML Schema document is available in physical level. Fully taken consideration of the features of the XML Schema, the structures of the logical UML class diagram are concise. Through a series of steps, the logical UML class diagram can be translated from the original conceptual UML class diagram, thereafter the XML Schema can be available in a one to one manner.
This paper presents a UML-based method for XML Schema design. The method includes three phases:conceptual, logical and physical level design. Standard UML class diagram is utilized in conceptual level, customized logical UML class diagram is utilized in logical level and textual XML Schema document is available in physical level. Fully taken consideration of the features of the XML Schema, the structures of the logical UML class diagram are concise. Through a series of steps, the logical UML class diagram can be translated from the original conceptual UML class diagram, thereafter the XML Schema can be available in a one to one manner.
2006, 35(3): 392-395.
Abstract:
The packet marking schemes of IP traceback are analyzed. A scheme based on probabilistic marking packet using algebraic coding theory is discussed. The victim can use the edges sampled in these packets to reconstruct attack path. The method of marking packets and reconstructing attack path algorithm are analyzed in this paper. This technique has very low requirements of network and router and supports incremental deployment in IPv6.
The packet marking schemes of IP traceback are analyzed. A scheme based on probabilistic marking packet using algebraic coding theory is discussed. The victim can use the edges sampled in these packets to reconstruct attack path. The method of marking packets and reconstructing attack path algorithm are analyzed in this paper. This technique has very low requirements of network and router and supports incremental deployment in IPv6.
2006, 35(3): 396-398.
Abstract:
Based on the study of object-oriented software measurement and the characteristics of object-oriented technology, some improvements on C&K metrics suite are discussed in this paper. Some new metrics for class size and class hierarchy measurement are presented including complexity per class, number of methods, depth of multi-inheritance tree and number of father etc. Meanwhile, the applications of these improved metrics in software engineering are discussed. All the improvements can supplement C&K metrics and optimize the result of measurement.
Based on the study of object-oriented software measurement and the characteristics of object-oriented technology, some improvements on C&K metrics suite are discussed in this paper. Some new metrics for class size and class hierarchy measurement are presented including complexity per class, number of methods, depth of multi-inheritance tree and number of father etc. Meanwhile, the applications of these improved metrics in software engineering are discussed. All the improvements can supplement C&K metrics and optimize the result of measurement.
2006, 35(3): 399-402.
Abstract:
A Generalized Shunting Inhibition Neuron (GSIN) model is proposed by analyzing shortcomings of the normal shunting neuron model. A new feed forward neural network architecture based on GSIN, naming Generalized Shunting Inhibition Neural Network (GSINN), and its learning algorithm are then introduced. Finally, the GSINN is applied to several benchmark classification problems, and their performance is compared with the performances of Shunting Inhibitory Artificial Neural Network (SIANN) and BP networks, and the effectiveness of the proposed network structure and learning algorithm is verified. Experimental results show that a single GSIN and simple GSINN can outperform both the SIANN and Back Propagation (BP) network.
A Generalized Shunting Inhibition Neuron (GSIN) model is proposed by analyzing shortcomings of the normal shunting neuron model. A new feed forward neural network architecture based on GSIN, naming Generalized Shunting Inhibition Neural Network (GSINN), and its learning algorithm are then introduced. Finally, the GSINN is applied to several benchmark classification problems, and their performance is compared with the performances of Shunting Inhibitory Artificial Neural Network (SIANN) and BP networks, and the effectiveness of the proposed network structure and learning algorithm is verified. Experimental results show that a single GSIN and simple GSINN can outperform both the SIANN and Back Propagation (BP) network.
2006, 35(3): 403-405.
Abstract:
This paper introduces the conception and common categories of Redundant Array of Inexpensive Disks (RAID). The reliability value for three RAIDs, RAID0, RAID3 and RAID10, is analyzed. The analysis results show that the value of the reliability for RAID base on standard model can be calculated, and RAID10 is better than other RAIDS on capability and reliability.
This paper introduces the conception and common categories of Redundant Array of Inexpensive Disks (RAID). The reliability value for three RAIDs, RAID0, RAID3 and RAID10, is analyzed. The analysis results show that the value of the reliability for RAID base on standard model can be calculated, and RAID10 is better than other RAIDS on capability and reliability.
2006, 35(3): 406-408.
Abstract:
This paper introduces a method for identifying flame based on net camera In this method the streaming media is compressed according to MPEG4 format. The image processing program divides the BMP images from streaming media, then identifies and positions the flame according to the characteristic in chroma and twinkling.
This paper introduces a method for identifying flame based on net camera In this method the streaming media is compressed according to MPEG4 format. The image processing program divides the BMP images from streaming media, then identifies and positions the flame according to the characteristic in chroma and twinkling.
2006, 35(3): 409-411,432.
Abstract:
Cobalt-doped SnO2 nano-powders (Sn1-xCoxO2) are fabricated by wet chemical method. It is found that cobalt exists as an ionic state instead of pure metal state. There is no other impurity phase in the samples except for rutile type SnO2. With the increase of cobalt concentration, the crystalline size decreases, the optical absorption spectra red-shiftes, and the band gap decreases linearly. The decrease of magnetic moment per cobalt atom with the increase of cobalt concentration could be attributed to the increased number of cobalt atoms occupying adjacent cation lattice positions that result in antiferromagnetic alignment.
Cobalt-doped SnO2 nano-powders (Sn1-xCoxO2) are fabricated by wet chemical method. It is found that cobalt exists as an ionic state instead of pure metal state. There is no other impurity phase in the samples except for rutile type SnO2. With the increase of cobalt concentration, the crystalline size decreases, the optical absorption spectra red-shiftes, and the band gap decreases linearly. The decrease of magnetic moment per cobalt atom with the increase of cobalt concentration could be attributed to the increased number of cobalt atoms occupying adjacent cation lattice positions that result in antiferromagnetic alignment.
2006, 35(3): 412-414,418.
Abstract:
In this paper, we present a method for designing multilayer-radar-absorbing material. Generally, low reflectance within low frequency band is difficultly obtained by using ordinary lossy material. Magnetic metal fiber may have larger permeability at low microwave band than ordinary lossy materials have, but until now, no design has been performed for absorbing material containing magnetic fiber. We design the absorbing material by using magnetic fiber composite as a layer together with several ordinary material layers. The parameters and thicknesses of these layers are optimized by a modified genetic algorithm. The obtained results show that the reflectance at low frequency have been improved considerably.
In this paper, we present a method for designing multilayer-radar-absorbing material. Generally, low reflectance within low frequency band is difficultly obtained by using ordinary lossy material. Magnetic metal fiber may have larger permeability at low microwave band than ordinary lossy materials have, but until now, no design has been performed for absorbing material containing magnetic fiber. We design the absorbing material by using magnetic fiber composite as a layer together with several ordinary material layers. The parameters and thicknesses of these layers are optimized by a modified genetic algorithm. The obtained results show that the reflectance at low frequency have been improved considerably.
2006, 35(3): 415-418.
Abstract:
Tissue's pathological changes are observed and analyzed and paraffin sections are studied by fourier transform infrared spectroscopy technology on lung tissues to observe the effluence on tissues of rats which are exposed to high intensity low-frequency electromagnetic fields.Results show that there are some differences between exposed group and control group on both lung tissue sections and spectra,These differences is are mainly caused by changes in composition and structure of micro-molecules and changes of vibrational modes of the function groups in biological tissues.Results indicate that high density low-frequency electromagnetic radiation may be caninogenic for lung tissues of rats with long time exposure.
Tissue's pathological changes are observed and analyzed and paraffin sections are studied by fourier transform infrared spectroscopy technology on lung tissues to observe the effluence on tissues of rats which are exposed to high intensity low-frequency electromagnetic fields.Results show that there are some differences between exposed group and control group on both lung tissue sections and spectra,These differences is are mainly caused by changes in composition and structure of micro-molecules and changes of vibrational modes of the function groups in biological tissues.Results indicate that high density low-frequency electromagnetic radiation may be caninogenic for lung tissues of rats with long time exposure.
2006, 35(3): 419-422.
Abstract:
The maximum order of A5's element is 5. By using Lagrange theorem, the could-be order of the element of A5's subgroup are 2 and 5, but in general the inverse of the Lagrange theorem is not hold, whether the ten-order subgroup of A5 exist is still a problem to discuss. Based on the researching and calculation on the power of 5-cyclic permutation, the forming law of the 10 order subgroup of A5 is fined and therefore 6 10-order subgroup of A5 are obtained by using the law.
The maximum order of A5's element is 5. By using Lagrange theorem, the could-be order of the element of A5's subgroup are 2 and 5, but in general the inverse of the Lagrange theorem is not hold, whether the ten-order subgroup of A5 exist is still a problem to discuss. Based on the researching and calculation on the power of 5-cyclic permutation, the forming law of the 10 order subgroup of A5 is fined and therefore 6 10-order subgroup of A5 are obtained by using the law.
2006, 35(3): 423-425.
Abstract:
Through using the reversed hazard rate of random variables, this paper derives an expression of Fisher information based on the knowledge of Fisher information in hand. The expression of the retained and lost Fisher information of a random variable due to censorship is produced. So the component of Fisher information under censorship is detected. Then, this paper gives a property of Fisher information in censored data.
Through using the reversed hazard rate of random variables, this paper derives an expression of Fisher information based on the knowledge of Fisher information in hand. The expression of the retained and lost Fisher information of a random variable due to censorship is produced. So the component of Fisher information under censorship is detected. Then, this paper gives a property of Fisher information in censored data.
2006, 35(3): 426-428.
Abstract:
Are studied by introducing the modified surplus and modified claim amounts,Some questions in a discrete risk model.The lundberg upper bound of the ultimate probability of ruin is obtained by using the martingale approach.
Are studied by introducing the modified surplus and modified claim amounts,Some questions in a discrete risk model.The lundberg upper bound of the ultimate probability of ruin is obtained by using the martingale approach.
2006, 35(3): 429-432.
Abstract:
From the point of competitive strategy, using price competition model, the paper analyzes debt strategically affecting on product price under the condition of uncertainty, builds up debt strategic effect model based on two kinds of classification and gets general methods of choosing debt level in different industries. A new way for firm's debt financing decision is presented.
From the point of competitive strategy, using price competition model, the paper analyzes debt strategically affecting on product price under the condition of uncertainty, builds up debt strategic effect model based on two kinds of classification and gets general methods of choosing debt level in different industries. A new way for firm's debt financing decision is presented.
