Email alert
2002 Vol. 24, No. 7
Display Method:
2002, 24(7): 865-871.
Abstract:
A novel method based on orthogonal redundancy sampling is proposed for har-monies estimation, which utilizes orthogonal redundancy sampling to obtain a signal with the information of its frequency range, and decomposes signal s auto-correlation matrix into the summation of rank-1 matrixes to estimate frequency of harmonics with the information. The results are compared with that of MUSIC and AR algorithms, and find that this method has a lower SNR threshold and higher range resolution.
A novel method based on orthogonal redundancy sampling is proposed for har-monies estimation, which utilizes orthogonal redundancy sampling to obtain a signal with the information of its frequency range, and decomposes signal s auto-correlation matrix into the summation of rank-1 matrixes to estimate frequency of harmonics with the information. The results are compared with that of MUSIC and AR algorithms, and find that this method has a lower SNR threshold and higher range resolution.
2002, 24(7): 872-878.
Abstract:
Under the condition of spatial subsampling, with flexible definition of fourth-order cumulant matrix of L-shaped array antennas output, utilizing the blind Gaussian and aperture extension properties of high-order cumulant, the estimation of frequency and two-dimensional(2D) Angles Of Arrivals (AOAs) are separated from the estimation of polarization of spatial non-Gaussian broad-band signals. An algorithm based on fourth-order cumulant ESPRIT and integral search is proposed to jointly estimate the parameters in multi-signal and arbitrary Gaussian noise environment. The array antennas are consisted of crossed dipoles that are parallel to the coordinate axes. Numerical simulation results verify the feasibility of this method.
Under the condition of spatial subsampling, with flexible definition of fourth-order cumulant matrix of L-shaped array antennas output, utilizing the blind Gaussian and aperture extension properties of high-order cumulant, the estimation of frequency and two-dimensional(2D) Angles Of Arrivals (AOAs) are separated from the estimation of polarization of spatial non-Gaussian broad-band signals. An algorithm based on fourth-order cumulant ESPRIT and integral search is proposed to jointly estimate the parameters in multi-signal and arbitrary Gaussian noise environment. The array antennas are consisted of crossed dipoles that are parallel to the coordinate axes. Numerical simulation results verify the feasibility of this method.
2002, 24(7): 879-886.
Abstract:
The key problem of eigen-subspace methods is the estimation of signal or noise subspace. In practical situations, there exist signals whose statistic characteristics always change over time. To obtain the real-time estimates of signal parameters, it is necessary to update the signal/noise subspace according to newly received array sampled output. In this paper, 3D Unitary ESPRIT algorithm is proposed to achieve the combined estimation of 2D DOA and carrier frequency of impinging signals, then a subspace tracking algorithm based on spherically averaged ULV decomposition is presented. With combination of the above subspace tracking algorithm with 3D Unitary ESPRIT algorithm, adaptive 3D Unitary ESPRIT algorithm is presented to track the time-varying multidimensional parameter estimates. Computer simulation results are provided to demonstrate the effectiveness of the proposed algorithm.
The key problem of eigen-subspace methods is the estimation of signal or noise subspace. In practical situations, there exist signals whose statistic characteristics always change over time. To obtain the real-time estimates of signal parameters, it is necessary to update the signal/noise subspace according to newly received array sampled output. In this paper, 3D Unitary ESPRIT algorithm is proposed to achieve the combined estimation of 2D DOA and carrier frequency of impinging signals, then a subspace tracking algorithm based on spherically averaged ULV decomposition is presented. With combination of the above subspace tracking algorithm with 3D Unitary ESPRIT algorithm, adaptive 3D Unitary ESPRIT algorithm is presented to track the time-varying multidimensional parameter estimates. Computer simulation results are provided to demonstrate the effectiveness of the proposed algorithm.
2002, 24(7): 887-893.
Abstract:
Cyclostationarity of AM signals is analyzed and time-difference extraction method based on Cycle Cross-Correlation(CCC) is presented in this paper. An Extended CCC method AM-ECCC method for moving targets is given also. Simulation results show that the given method can work more efficiently in highly corruptive environment than correlation method.
Cyclostationarity of AM signals is analyzed and time-difference extraction method based on Cycle Cross-Correlation(CCC) is presented in this paper. An Extended CCC method AM-ECCC method for moving targets is given also. Simulation results show that the given method can work more efficiently in highly corruptive environment than correlation method.
2002, 24(7): 894-899.
Abstract:
A modified DS evidence combination method based on evidence classification is introduced. Such a method imitates human s behavior in classification to solve the problem of combing evidences with high degree of conflict. When a sequence of beliefs is available, first, those beliefs are classified into several sets to decide which ones correspond with each other and can be combined with DS rule reasonably, and then all the combination results are integrated with weight rule to form the final result.
A modified DS evidence combination method based on evidence classification is introduced. Such a method imitates human s behavior in classification to solve the problem of combing evidences with high degree of conflict. When a sequence of beliefs is available, first, those beliefs are classified into several sets to decide which ones correspond with each other and can be combined with DS rule reasonably, and then all the combination results are integrated with weight rule to form the final result.
2002, 24(7): 900-904.
Abstract:
A novel wavelet-based vector quantization (VQ) algorithm is presented in this paper. In this algorithm, first, 9/7 biorthogonal wavelet transform decomposes the original image, then the correlation of the three directions wavelet coefficients is used to construct the band-cross vector. Therefore, the high coding efficiercy is obtained. In addition, a new VQ technology is used in this algorithm to improve the coding efficiency and the quality of the code-book. The experimental results show that this coding scheme, no entropy-coding, can get a reconstruction quality(PSNR32dB) at a very low bit rate(0.142 bpp) with high feasibility.
A novel wavelet-based vector quantization (VQ) algorithm is presented in this paper. In this algorithm, first, 9/7 biorthogonal wavelet transform decomposes the original image, then the correlation of the three directions wavelet coefficients is used to construct the band-cross vector. Therefore, the high coding efficiercy is obtained. In addition, a new VQ technology is used in this algorithm to improve the coding efficiency and the quality of the code-book. The experimental results show that this coding scheme, no entropy-coding, can get a reconstruction quality(PSNR32dB) at a very low bit rate(0.142 bpp) with high feasibility.
2002, 24(7): 905-909.
Abstract:
In this paper, an automatic algorithm of image enhancement based on fuzzy sets is proposed by defining a new membership function. This algorithm is performed through extracting the fuzzy property plane from the spatial domain firstly, executing the fuzzy enhancement, and obtaining the modified spatial domain image finally. In order to get much better quality of the image the contrast among the regions is increased using the fuzzy enhancement. This algorithm is also extended to the enhancement of the multi-threshold image.
In this paper, an automatic algorithm of image enhancement based on fuzzy sets is proposed by defining a new membership function. This algorithm is performed through extracting the fuzzy property plane from the spatial domain firstly, executing the fuzzy enhancement, and obtaining the modified spatial domain image finally. In order to get much better quality of the image the contrast among the regions is increased using the fuzzy enhancement. This algorithm is also extended to the enhancement of the multi-threshold image.
2002, 24(7): 910-915.
Abstract:
Euclidean algorithm can be used to design Perfect Reconstruction (PR) filter banks. An algorithm of Euclidean polynomial factorization and the theory of two-channel filter banks with low delay are discussed in this paper. A major problem of Euclidean algorithm with the polynomial case is that the solution is not unique. To overcome this problem, a proposition about factorization method for solving this problem has been given and proven. A highly effective method based on the factorization for designing two-channel PR filter banks satisfying low-delay is presented. Final, some examples are given to illustrate that the method is effective.
Euclidean algorithm can be used to design Perfect Reconstruction (PR) filter banks. An algorithm of Euclidean polynomial factorization and the theory of two-channel filter banks with low delay are discussed in this paper. A major problem of Euclidean algorithm with the polynomial case is that the solution is not unique. To overcome this problem, a proposition about factorization method for solving this problem has been given and proven. A highly effective method based on the factorization for designing two-channel PR filter banks satisfying low-delay is presented. Final, some examples are given to illustrate that the method is effective.
2002, 24(7): 916-920.
Abstract:
An Artificial Neural Network (ANN) is proposed which is implemented through Pulse Width Modulation (PWM) technique. The weight of the neural network is stored in digital memory, while the synapse circuits are composed of switches. The capacitor integral circuits and the current sample/hold circuits are used to implement the Weighted summation Operation (WSO) that are required in the realization of a general ANN. This circuit is easy to be integrated because of its simple structure, which could be used to construct the feed forward neural networks, for example, to implement ANN for an XOR function.
An Artificial Neural Network (ANN) is proposed which is implemented through Pulse Width Modulation (PWM) technique. The weight of the neural network is stored in digital memory, while the synapse circuits are composed of switches. The capacitor integral circuits and the current sample/hold circuits are used to implement the Weighted summation Operation (WSO) that are required in the realization of a general ANN. This circuit is easy to be integrated because of its simple structure, which could be used to construct the feed forward neural networks, for example, to implement ANN for an XOR function.
2002, 24(7): 921-928.
Abstract:
In this paper, the characteristics of B-spline basis functions are first introduced, and the universal approximation properties of hierarchical fuzzy systems based on B-spline basis functions are verified, and then, another new proof on approximation property of multiwavelet neural network is given using the equivalence relation between B-spline-basis-function-based fuzzy system(HBFS) and wavelet neural network. All of these provide solid theoretical foundation for HBFS s application.
In this paper, the characteristics of B-spline basis functions are first introduced, and the universal approximation properties of hierarchical fuzzy systems based on B-spline basis functions are verified, and then, another new proof on approximation property of multiwavelet neural network is given using the equivalence relation between B-spline-basis-function-based fuzzy system(HBFS) and wavelet neural network. All of these provide solid theoretical foundation for HBFS s application.
2002, 24(7): 929-934.
Abstract:
The problem of detecting targets in the cluttered circumstance with wide-band high-resolution radar is studied. A way of improving SCR(Signal-to-Clutter-Ratio) is cited by linear integration filtering in polarization domain. With the intrinsic high resolution of wide-band waveform, one-dimension range profile of radar target can be obtained. In connection with the fact that the values of Stokes vector and polarization attitude of target greatly differ from which of cluttered circumstance, a new target detection method is presented. Simulation experimental results show that radars detection performance can be improved significantly by using the algorithm presented in this paper.
The problem of detecting targets in the cluttered circumstance with wide-band high-resolution radar is studied. A way of improving SCR(Signal-to-Clutter-Ratio) is cited by linear integration filtering in polarization domain. With the intrinsic high resolution of wide-band waveform, one-dimension range profile of radar target can be obtained. In connection with the fact that the values of Stokes vector and polarization attitude of target greatly differ from which of cluttered circumstance, a new target detection method is presented. Simulation experimental results show that radars detection performance can be improved significantly by using the algorithm presented in this paper.
2002, 24(7): 935-943.
Abstract:
Based on the principles of the space-borne ScanSAR, some important parameters to define the position of sub-swathes, the timing sequence of the system, the choice of the PRFs, the number of the samples of each burst and the SNR are investigated. Some critical problems about ScanSAR system are discussed, such as the number of the multi-looks, the azimuth resolution and the range ambiguities. A new concept of the range ambiguities is put forward, analyzed and verified. Based on above discussion and calculation, the data format for the practical ScanSAR system is suggested.
Based on the principles of the space-borne ScanSAR, some important parameters to define the position of sub-swathes, the timing sequence of the system, the choice of the PRFs, the number of the samples of each burst and the SNR are investigated. Some critical problems about ScanSAR system are discussed, such as the number of the multi-looks, the azimuth resolution and the range ambiguities. A new concept of the range ambiguities is put forward, analyzed and verified. Based on above discussion and calculation, the data format for the practical ScanSAR system is suggested.
2002, 24(7): 944-947.
Abstract:
In this paper a mandarin continuous speech recognition system is implemented based on classical Hidden Markov Model (HMM). The performance of classical HMM is quantitatively analyzed and evaluated.
In this paper a mandarin continuous speech recognition system is implemented based on classical Hidden Markov Model (HMM). The performance of classical HMM is quantitatively analyzed and evaluated.
2002, 24(7): 948-953.
Abstract:
In multimedia communication networks, the multicasting is increasingly important. Nodes in the networks will likely vary in their abilities to support multicasting. Some nodes do not have multicasting ability, other nodes may be limited in the number of multicast copies they can reasonably make. So in this paper the degree-constraint of the node is used to represent its multicasting ability. Furthermore many services in networks require that the accumulated delay from the source node to every destination node can not exceed a pre-specified delay bound. In this paper the multicast routing algorithm with degree and delay constraints is proposed, and the Lagrange relaxation method is used to solve this problem.
In multimedia communication networks, the multicasting is increasingly important. Nodes in the networks will likely vary in their abilities to support multicasting. Some nodes do not have multicasting ability, other nodes may be limited in the number of multicast copies they can reasonably make. So in this paper the degree-constraint of the node is used to represent its multicasting ability. Furthermore many services in networks require that the accumulated delay from the source node to every destination node can not exceed a pre-specified delay bound. In this paper the multicast routing algorithm with degree and delay constraints is proposed, and the Lagrange relaxation method is used to solve this problem.
2002, 24(7): 954-960.
Abstract:
The scheduling algorithms of routers in the Internet are very important to the QoS (Quality of Service) which can be derived by users flows. A new scheduling algorithmLogarithmic Proportional Fair queueing (LPFQ) scheduling algorithm is presented, and its characteristics are discussed using theoretical analysis and simulations. Theoretical analysis and simulation results demonstrate that LPFQ can not only achieve the compromise among fairness, network utility and total delay, but also punish malicious flows and encourage users to use proper end-to-end flow control algorithms.
The scheduling algorithms of routers in the Internet are very important to the QoS (Quality of Service) which can be derived by users flows. A new scheduling algorithmLogarithmic Proportional Fair queueing (LPFQ) scheduling algorithm is presented, and its characteristics are discussed using theoretical analysis and simulations. Theoretical analysis and simulation results demonstrate that LPFQ can not only achieve the compromise among fairness, network utility and total delay, but also punish malicious flows and encourage users to use proper end-to-end flow control algorithms.
2002, 24(7): 961-967.
Abstract:
In this paper, the approach to produce the complex wavelet packet basis function is given. A complex wavelet packet function based frequency-hopping multiple access communication system, which employs antenna diversity and decision feedback equalization, is proposed. The simulation results show that the new multiple access system outperforms the analogy system based on real wavelet packet basis function modulation. In addition, it has high spectrum efficiency, high security and good narrow band interference rejection capability. Furthermore, the developed system is also easy to support multirate voice and date services.
In this paper, the approach to produce the complex wavelet packet basis function is given. A complex wavelet packet function based frequency-hopping multiple access communication system, which employs antenna diversity and decision feedback equalization, is proposed. The simulation results show that the new multiple access system outperforms the analogy system based on real wavelet packet basis function modulation. In addition, it has high spectrum efficiency, high security and good narrow band interference rejection capability. Furthermore, the developed system is also easy to support multirate voice and date services.
2002, 24(7): 968-975.
Abstract:
In order to analyze the subharmonic load pull characteristic of MESFET power amplifier driven by CDMA type wide band signal, a set of Volterra transfer funcion expressions are deducted with the MESFET nonlinear model. With these expressions, the relation between the IM3 (third order intermodulation) and the subharmonic impedance is established, and the characteristic of the subharmonic load pull is analyzed. the expression of the optimal subharmonic load is also given with the analysis. It is proved in the paper that the reactance part of the subharmonic impedance is the main factor that leads to the difference of the two IM3. The result of the analysis agrees well with the simulation result got from ADS with harmonic balance method.
In order to analyze the subharmonic load pull characteristic of MESFET power amplifier driven by CDMA type wide band signal, a set of Volterra transfer funcion expressions are deducted with the MESFET nonlinear model. With these expressions, the relation between the IM3 (third order intermodulation) and the subharmonic impedance is established, and the characteristic of the subharmonic load pull is analyzed. the expression of the optimal subharmonic load is also given with the analysis. It is proved in the paper that the reactance part of the subharmonic impedance is the main factor that leads to the difference of the two IM3. The result of the analysis agrees well with the simulation result got from ADS with harmonic balance method.
2002, 24(7): 976-981.
Abstract:
Analytical solutions of electromagnetic fields in the elliptical chirowaveguide have been studied. The model fields distribution and modal characteristic equations are obtained. The numerical solutions for the modal characteristic equations of low-order modes have been investigated. Effects of chirality parameters on the characteristics such as dispersion curves and cutoff wavelengths of guided modes in the elliptical chirowaveguide are presented. The first-order mode will be changed as the chirality parameter increased.
Analytical solutions of electromagnetic fields in the elliptical chirowaveguide have been studied. The model fields distribution and modal characteristic equations are obtained. The numerical solutions for the modal characteristic equations of low-order modes have been investigated. Effects of chirality parameters on the characteristics such as dispersion curves and cutoff wavelengths of guided modes in the elliptical chirowaveguide are presented. The first-order mode will be changed as the chirality parameter increased.
2002, 24(7): 982-986.
Abstract:
The roles of hot electrons and holes in dielectric breakdown of the thin gate oxide have been quantitatively investigated in this paper, The changes of threshold voltage have been discussed under different stress conditions. This paper is the first report that points out the cooperation of hot electrons and holes is essential for the time dependent dielectric breakdown in thin gate oxide. A detailed theory analysis is made and a two-step models of thin gate oxide is presented. The first step is injected hot electrons create trap centers in thin gate oxide, and the second step is thin gate oxide breakdown induced by hole trapping.
The roles of hot electrons and holes in dielectric breakdown of the thin gate oxide have been quantitatively investigated in this paper, The changes of threshold voltage have been discussed under different stress conditions. This paper is the first report that points out the cooperation of hot electrons and holes is essential for the time dependent dielectric breakdown in thin gate oxide. A detailed theory analysis is made and a two-step models of thin gate oxide is presented. The first step is injected hot electrons create trap centers in thin gate oxide, and the second step is thin gate oxide breakdown induced by hole trapping.
2002, 24(7): 987-991.
Abstract:
A new approach for continuous speech recognition based on neurocomputer CASS-ANDRA-I and syntactic analysis based on key words is proposed. By using CASSANDRA-I, the speed of the recognition is so rapid that recognition can be at real time. By using syntactic analysis based on key words, the performance of recognition improves greatly.
A new approach for continuous speech recognition based on neurocomputer CASS-ANDRA-I and syntactic analysis based on key words is proposed. By using CASSANDRA-I, the speed of the recognition is so rapid that recognition can be at real time. By using syntactic analysis based on key words, the performance of recognition improves greatly.
2002, 24(7): 992-996.
Abstract:
In this paper, a new method for time-varying ARM A model is introduced. It includes two procedures. First, using one method of signal decomposition, a signal is decomposed into some basic components; second, time-varying ARMA model is established in any of these basic components, and gets their time-frequency spectrum. This method can analyze complicated non-stationary nonlinear signal.
In this paper, a new method for time-varying ARM A model is introduced. It includes two procedures. First, using one method of signal decomposition, a signal is decomposed into some basic components; second, time-varying ARMA model is established in any of these basic components, and gets their time-frequency spectrum. This method can analyze complicated non-stationary nonlinear signal.
2002, 24(7): 997-999.
Abstract:
An improved protocol of electrical auction with sealed bids is presented by using nested interval theorem, bit commitment protocol and millionaire protocol. With the advantages of anonymity, privacy, efficiency, the improved protocol supports the second-price principle and optimal distribution of goods.
An improved protocol of electrical auction with sealed bids is presented by using nested interval theorem, bit commitment protocol and millionaire protocol. With the advantages of anonymity, privacy, efficiency, the improved protocol supports the second-price principle and optimal distribution of goods.
2002, 24(7): 1000-1004.
Abstract:
In this paper, the operating principle of UWB spread spectrum communication technique is described first, then the receiver signal processing is analyzed, the multiple access performance: BER and the number of users that system can support is evaluated. The numerical results is presented. These studies indicate that UWB has potential capability in the field of high speed wireless access in dense multipath environments.
In this paper, the operating principle of UWB spread spectrum communication technique is described first, then the receiver signal processing is analyzed, the multiple access performance: BER and the number of users that system can support is evaluated. The numerical results is presented. These studies indicate that UWB has potential capability in the field of high speed wireless access in dense multipath environments.
2002, 24(7): 1005-1008.
Abstract:
Hf film is coated on the surface of Mo-grids by vacuum ion beam aided deposition technology. The SEM, XRD and RBS analyses are performed on modified Mo-grids. Life-span tests of this modified Mo-grid are carried out by diode experiment. When the temperature of the modified Mo-grids is kept at 650 ℃, the modified Mo-grids electron emission can not be detected with microammeter after 1000h life-span. The experimental results show that after the grid is contaminated by electron emission substance Ba of the cathode Hf formed compounds restrain effectively electron emission of grid.
Hf film is coated on the surface of Mo-grids by vacuum ion beam aided deposition technology. The SEM, XRD and RBS analyses are performed on modified Mo-grids. Life-span tests of this modified Mo-grid are carried out by diode experiment. When the temperature of the modified Mo-grids is kept at 650 ℃, the modified Mo-grids electron emission can not be detected with microammeter after 1000h life-span. The experimental results show that after the grid is contaminated by electron emission substance Ba of the cathode Hf formed compounds restrain effectively electron emission of grid.
