Email alert
2007 Vol. 29, No. 9
Display Method:
2007, 29(9): 2033-2037.
doi: 10.3724/SP.J.1146.2006.00993
Abstract:
To achieve infrared image denoising and edge enhancement, a method based on wavelet forward and backward diffusion is introduced. Wavelet forward and backward diffusion is based on wavelet diffusion theory. It not only inherites iterative noise reduction and edge preserving features from wavelet diffusion, but also enhancemented image edge at the same time. In order to solve the problem of using wavelet modulus, it uses both wavelet modulus and local regularity to get elementary edge map, then get accurate edge map with geometric consistency and redesign wavelet forward and backward diffusion equation. Experiment shows that the method can effectively realize infrared image denoising and edge enhancement.
To achieve infrared image denoising and edge enhancement, a method based on wavelet forward and backward diffusion is introduced. Wavelet forward and backward diffusion is based on wavelet diffusion theory. It not only inherites iterative noise reduction and edge preserving features from wavelet diffusion, but also enhancemented image edge at the same time. In order to solve the problem of using wavelet modulus, it uses both wavelet modulus and local regularity to get elementary edge map, then get accurate edge map with geometric consistency and redesign wavelet forward and backward diffusion equation. Experiment shows that the method can effectively realize infrared image denoising and edge enhancement.
2007, 29(9): 2038-2041.
doi: 10.3724/SP.J.1146.2006.00226
Abstract:
Based on the fractional lower order theory a novel fractional time delay estimation algorithm referred to as EFMML (Explicit Fractional lower order Mix Modulated Lagrange) under stable distributed noise environment is proposed. The convergence condition of the proposed method is analyzed in theory. Computer simulation shows that the proposed method is robust for both Gaussian and non-Gaussian impulsive noises environments. And it is more effective than FSETDE (Fractional lower order Simplified Explicit Time Delay Estimation), an existing time delay estimation algorithm under stable distributed noise environment.
Based on the fractional lower order theory a novel fractional time delay estimation algorithm referred to as EFMML (Explicit Fractional lower order Mix Modulated Lagrange) under stable distributed noise environment is proposed. The convergence condition of the proposed method is analyzed in theory. Computer simulation shows that the proposed method is robust for both Gaussian and non-Gaussian impulsive noises environments. And it is more effective than FSETDE (Fractional lower order Simplified Explicit Time Delay Estimation), an existing time delay estimation algorithm under stable distributed noise environment.
2007, 29(9): 2042-2046.
doi: 10.3724/SP.J.1146.2006.00051
Abstract:
In this paper, a low complexity ESPRIT algorithm based on power method and QR decomposition is presented for direction finding, which doesnot require the priori knowledge of sources number and the predetermined threshold in separation of the signal and noise eigen-values. Firstly, the estimation of noise subspace is obtained by the power of covariance matrix and a novel source number detection method without eigen-decomposition is proposed based on QR decomposition. Furthermore, the eigen-vectors of signal subspace can be determined according to Q matrix, and then the directions of signals could be computed by the ESPRIT algorithm. In determining the source-number and subspace, the proposed algorithm has a substantial computational saving with the approximation performance compared with the Single-Vector-Decomposition (SVD) based algorithm. The simulation results demonstrate its effectiveness and robustness.
In this paper, a low complexity ESPRIT algorithm based on power method and QR decomposition is presented for direction finding, which doesnot require the priori knowledge of sources number and the predetermined threshold in separation of the signal and noise eigen-values. Firstly, the estimation of noise subspace is obtained by the power of covariance matrix and a novel source number detection method without eigen-decomposition is proposed based on QR decomposition. Furthermore, the eigen-vectors of signal subspace can be determined according to Q matrix, and then the directions of signals could be computed by the ESPRIT algorithm. In determining the source-number and subspace, the proposed algorithm has a substantial computational saving with the approximation performance compared with the Single-Vector-Decomposition (SVD) based algorithm. The simulation results demonstrate its effectiveness and robustness.
2007, 29(9): 2047-2049.
doi: 10.3724/SP.J.1146.2006.00319
Abstract:
Extracting the most discriminant low-dimensional face feature is an extremely critical step in Face Recognition (FR) systems. Linear Discriminant Analysis (LDA) is one of the most popular linear classification techniques for feature extraction. An optimized LDA algorithm is introduced to overcome questions existing in the traditional LDA algorithm for FR in this paper. The between-class scatter matrix is redefined in order to make the traditional Fisher criterion optimal and eliminate the effect that the edge of class has on selecting the optimal projection; at the same time, it avoids computing the inverse of matrix by means of factorization, and solves the Small Sample Size (SSS) problem. Adopting experiential method, the appropriate value of e is found, and then the optimal effect of face recognition is got. Experimental results show the recognition rate of this method is superior to the traditional LDA.
Extracting the most discriminant low-dimensional face feature is an extremely critical step in Face Recognition (FR) systems. Linear Discriminant Analysis (LDA) is one of the most popular linear classification techniques for feature extraction. An optimized LDA algorithm is introduced to overcome questions existing in the traditional LDA algorithm for FR in this paper. The between-class scatter matrix is redefined in order to make the traditional Fisher criterion optimal and eliminate the effect that the edge of class has on selecting the optimal projection; at the same time, it avoids computing the inverse of matrix by means of factorization, and solves the Small Sample Size (SSS) problem. Adopting experiential method, the appropriate value of e is found, and then the optimal effect of face recognition is got. Experimental results show the recognition rate of this method is superior to the traditional LDA.
2007, 29(9): 2050-2053.
doi: 10.3724/SP.J.1146.2006.00200
Abstract:
A Transcendental Information Based (TIB) triphone model tying structure adaptation strategy is delivered, and this strategy can improve the triphone model tying structure to suit the target co-pronunciation features with small amount of adaptation data. The TIB triphone model tying structure adaptation strategy uses the baseline acoustic models triphone model tying result as the transcendental model clustering center, with the adaptation data alignment by the baseline acoustic model, re-estimate the TIB triphone model clustering center and model tying structure recursively under maximum likelihood principle. The experiments show that the TIB triphone model tying structure adaptation strategy can improve the triphone model tying structure with only 2 hours adaptation corpus, and in the experiment of English acoustic model for Chinese speakers, the TIB strategy will increase the recognition accuracy rate from 74.59% to 83.63%.
A Transcendental Information Based (TIB) triphone model tying structure adaptation strategy is delivered, and this strategy can improve the triphone model tying structure to suit the target co-pronunciation features with small amount of adaptation data. The TIB triphone model tying structure adaptation strategy uses the baseline acoustic models triphone model tying result as the transcendental model clustering center, with the adaptation data alignment by the baseline acoustic model, re-estimate the TIB triphone model clustering center and model tying structure recursively under maximum likelihood principle. The experiments show that the TIB triphone model tying structure adaptation strategy can improve the triphone model tying structure with only 2 hours adaptation corpus, and in the experiment of English acoustic model for Chinese speakers, the TIB strategy will increase the recognition accuracy rate from 74.59% to 83.63%.
2007, 29(9): 2054-2057.
doi: 10.3724/SP.J.1146.2006.00164
Abstract:
This paper proposes a new speech enhancement algorithm to improve the quality of speech in low SNR environment. The algorithm estimates the noise spectrum in the form of prior SNR according to its Gaussian statistical model firstly. Then it estimates the prior SNR of each frequency bin utilizing the intra-frame as well as inter-frame smoothness to trace the quick change of the prior SNR. It also introduces a simple method to compute the speech presence probability to each frequency bin and obtain the new speech enhancement algorithm. Objective measurement combined with informal listening test shows that the new algorithm is more effective in eliminating musical noise and improving SNR obviously without significantly impairing the intelligibility of speech.
This paper proposes a new speech enhancement algorithm to improve the quality of speech in low SNR environment. The algorithm estimates the noise spectrum in the form of prior SNR according to its Gaussian statistical model firstly. Then it estimates the prior SNR of each frequency bin utilizing the intra-frame as well as inter-frame smoothness to trace the quick change of the prior SNR. It also introduces a simple method to compute the speech presence probability to each frequency bin and obtain the new speech enhancement algorithm. Objective measurement combined with informal listening test shows that the new algorithm is more effective in eliminating musical noise and improving SNR obviously without significantly impairing the intelligibility of speech.
2007, 29(9): 2058-2062.
doi: 10.3724/SP.J.1146.2006.00806
Abstract:
Machine intelligence requires convenient and effective estimation mode for sound localization. According to the tight relation of Head Related Transfer Function (HRTF) and location, this paper studies a sound localization method based on HRTF. Using the forward neural network, the direction information is extracted and multi coded in parallel, and then is weighted to estimate the sound position. Simulations show that the method is more accurate and compact. Finally, robust analysis is given to prove validity of simulation results.
Machine intelligence requires convenient and effective estimation mode for sound localization. According to the tight relation of Head Related Transfer Function (HRTF) and location, this paper studies a sound localization method based on HRTF. Using the forward neural network, the direction information is extracted and multi coded in parallel, and then is weighted to estimate the sound position. Simulations show that the method is more accurate and compact. Finally, robust analysis is given to prove validity of simulation results.
2007, 29(9): 2063-2066.
doi: 10.3724/SP.J.1146.2006.00222
Abstract:
Phoneme lattice based Confidence Measure (CM) is proposed in this paper. It makes use of phoneme lattices generated by a phoneme recognizer. Acoustic Model (AM) based CM is also introduced. For a decoded speech frame aligned to an HMM state, the CM based on AM is calculated. These two confidence measures are combined using a weighting factor to obtain a hybrid CM as they had different dynamic scales. On spontaneous conversational telephone database, the Figure Of Merit (FOM) achieves 17.0% relative improvement comparing to AM based CM.
Phoneme lattice based Confidence Measure (CM) is proposed in this paper. It makes use of phoneme lattices generated by a phoneme recognizer. Acoustic Model (AM) based CM is also introduced. For a decoded speech frame aligned to an HMM state, the CM based on AM is calculated. These two confidence measures are combined using a weighting factor to obtain a hybrid CM as they had different dynamic scales. On spontaneous conversational telephone database, the Figure Of Merit (FOM) achieves 17.0% relative improvement comparing to AM based CM.
2007, 29(9): 2067-2072.
doi: 10.3724/SP.J.1146.2006.00120
Abstract:
With fast development and boosting of stream media applications, automatic classification of audio signals becomes one of the hotspots on research and engineering. Since mood classification of music is involved with integrated representation and classification of social and natural properties of music, mechanism selection and architecture optimization should be implemented on the basis of different traditional music representations and classification methods. This paper discusses formation of weak classifiers in AdaBoost algorithm based on K-L transformation and GMM training and realizes mood classification of music with multi-layer classifier architecture. The experiments classify 163 songs into four mood classes: calm, sad, exciting and pleasant with 97.5% accuracy on training data and 93.9% accuracy on test data, which proves feasibility and potential value of this method.
With fast development and boosting of stream media applications, automatic classification of audio signals becomes one of the hotspots on research and engineering. Since mood classification of music is involved with integrated representation and classification of social and natural properties of music, mechanism selection and architecture optimization should be implemented on the basis of different traditional music representations and classification methods. This paper discusses formation of weak classifiers in AdaBoost algorithm based on K-L transformation and GMM training and realizes mood classification of music with multi-layer classifier architecture. The experiments classify 163 songs into four mood classes: calm, sad, exciting and pleasant with 97.5% accuracy on training data and 93.9% accuracy on test data, which proves feasibility and potential value of this method.
2007, 29(9): 2073-2076.
doi: 10.3724/SP.J.1146.2006.00188
Abstract:
Kalman filter is a recursive filtering method based on minimum variance estimation, but it assumes that the signals state model is exactly known, which restricts its application in practice. Nevertheless the systems state equation can be obtained through the identification of systems by using neural networks good abilities of non-linear mapping. In contrast to some classic improved algorithm of Kalman filtering, this method has the advantages of wide application range, simple and feasible mathematical modeling. In this paper, the method which integrates neural network and Kalman filter is implemented for image restoration. The experiment result shows that the provided method is effective and available.
Kalman filter is a recursive filtering method based on minimum variance estimation, but it assumes that the signals state model is exactly known, which restricts its application in practice. Nevertheless the systems state equation can be obtained through the identification of systems by using neural networks good abilities of non-linear mapping. In contrast to some classic improved algorithm of Kalman filtering, this method has the advantages of wide application range, simple and feasible mathematical modeling. In this paper, the method which integrates neural network and Kalman filter is implemented for image restoration. The experiment result shows that the provided method is effective and available.
2007, 29(9): 2077-2081.
doi: 10.3724/SP.J.1146.2006.00055
Abstract:
To the questions of Threat Assessment (TA), a technique for TA based on intuitionistic fuzzy reasoning is proposed. First, the major factors of effects of attacking targets from space or air on TA in joint air defense operations, nondeterminacy to TA, and the properties and vulnerabilities of the existing TA methods are analyzed. A model for TA measurement rank of threat degree is exposed. Then, the membership and nonmembership functions for input/output state variables are devised. The methods for quantification and measurement transformation to input variables in fuzzy strategy are addressed. The inference rules of the system are constructed with algorithms for reasoning and defuzzification devised. Subsequently, completeness and interactivity and consistency of rules contained in the rulebase are checked with a verification method to the rulebase presented. Finally, the validity is checked by providing TA instances with 20 typical targets.
To the questions of Threat Assessment (TA), a technique for TA based on intuitionistic fuzzy reasoning is proposed. First, the major factors of effects of attacking targets from space or air on TA in joint air defense operations, nondeterminacy to TA, and the properties and vulnerabilities of the existing TA methods are analyzed. A model for TA measurement rank of threat degree is exposed. Then, the membership and nonmembership functions for input/output state variables are devised. The methods for quantification and measurement transformation to input variables in fuzzy strategy are addressed. The inference rules of the system are constructed with algorithms for reasoning and defuzzification devised. Subsequently, completeness and interactivity and consistency of rules contained in the rulebase are checked with a verification method to the rulebase presented. Finally, the validity is checked by providing TA instances with 20 typical targets.
2007, 29(9): 2082-2085.
doi: 10.3724/SP.J.1146.2006.00119
Abstract:
In order to process large-scale real text, a method of building Chinese parser based on lexicalized model is proposed. First, a unified approach for segmentation and part of speech tagging is proposed based on hidden Markov model. The method not only conservers the merits of HMM which is simple and efficient but also improves the tagging accuracy. Then the head-driven model is used to recognize phrases. Head-driven model is a well-known English parsing model; we combine it with segmentation and POS tagging model and thus build a Chinese parser that can operate at the character level. The parser is evaluated on the standard test set. It achieves 77.57% precision and 74.96% recall and outperforms the only previous comparable work significantly.
In order to process large-scale real text, a method of building Chinese parser based on lexicalized model is proposed. First, a unified approach for segmentation and part of speech tagging is proposed based on hidden Markov model. The method not only conservers the merits of HMM which is simple and efficient but also improves the tagging accuracy. Then the head-driven model is used to recognize phrases. Head-driven model is a well-known English parsing model; we combine it with segmentation and POS tagging model and thus build a Chinese parser that can operate at the character level. The parser is evaluated on the standard test set. It achieves 77.57% precision and 74.96% recall and outperforms the only previous comparable work significantly.
2007, 29(9): 2086-2089.
doi: 10.3724/SP.J.1146.2006.00185
Abstract:
The theory and its implementation on the adaptive window length in the analysis of time frequency are presented in this paper. This method is used to analysis the in-pulse characteristics of pseudo-code fuse signals such as Pseudo-Code Phase Modulation (PCPM), pseudo-code phase modulation combined by Sinusoid Frequency Modulation (SFM) and pseudo-code phase modulation combined by Linear Frequency Modulation (LFM). The simulation result shows that both the characteristic of carry frequency modulation and phase changed position caused by the pseudo-code can be extracted effectively in SNR=10dB.
The theory and its implementation on the adaptive window length in the analysis of time frequency are presented in this paper. This method is used to analysis the in-pulse characteristics of pseudo-code fuse signals such as Pseudo-Code Phase Modulation (PCPM), pseudo-code phase modulation combined by Sinusoid Frequency Modulation (SFM) and pseudo-code phase modulation combined by Linear Frequency Modulation (LFM). The simulation result shows that both the characteristic of carry frequency modulation and phase changed position caused by the pseudo-code can be extracted effectively in SNR=10dB.
2007, 29(9): 2090-2093.
doi: 10.3724/SP.J.1146.2006.00201
Abstract:
This paper firstly studies the characteristic of the UWB radar frequency response and presents a parameterized method to extract the poles of UWB still target. This method regards the high frequency echo as an ARMA model and poles of this ARMA model express the scattering type and range of the scatter. Secondly, according the actual situation, the motive target model of UWB radar is presented. This model is composed of a still model and a polynomial phase part which is caused by the motion of target. Based this character, a cyclostationary method can be used to estimate the motion parameters of the target. After the motive target model compensated by the estimated parameters, the eigenvalues can be extracted by the state space method. Simulation shows 1-D range and scattering mechanism can be correctly obtained and proves the validity of the method.
This paper firstly studies the characteristic of the UWB radar frequency response and presents a parameterized method to extract the poles of UWB still target. This method regards the high frequency echo as an ARMA model and poles of this ARMA model express the scattering type and range of the scatter. Secondly, according the actual situation, the motive target model of UWB radar is presented. This model is composed of a still model and a polynomial phase part which is caused by the motion of target. Based this character, a cyclostationary method can be used to estimate the motion parameters of the target. After the motive target model compensated by the estimated parameters, the eigenvalues can be extracted by the state space method. Simulation shows 1-D range and scattering mechanism can be correctly obtained and proves the validity of the method.
2007, 29(9): 2094-2097.
doi: 10.3724/SP.J.1146.2006.00129
Abstract:
The bistatic SAR overall slant range history of target response is essentially different from the monostatic slant range history; hence, bistatic SAR imaging processing can not be achieved by purely monostatic approaches. To address the problem, a wavenumber domain algorithm is derived, which is based on the one-order approximation of the point scatter responses two-dimensional spectrum phase. Bistatic SAR data of simulated point and surface scatter response are focused by the algorithm and the quality of the image demonstrates the effectiveness and high resolution of the algorithm in large squint angle and large base line length case.
The bistatic SAR overall slant range history of target response is essentially different from the monostatic slant range history; hence, bistatic SAR imaging processing can not be achieved by purely monostatic approaches. To address the problem, a wavenumber domain algorithm is derived, which is based on the one-order approximation of the point scatter responses two-dimensional spectrum phase. Bistatic SAR data of simulated point and surface scatter response are focused by the algorithm and the quality of the image demonstrates the effectiveness and high resolution of the algorithm in large squint angle and large base line length case.
2007, 29(9): 2098-2100.
doi: 10.3724/SP.J.1146.2006.00208
Abstract:
To the big squint mode of bistatic SAR system，based on the echo model of bistatic SAR and range equation in RD area, properly to with the NCS algorithm of conventional SAR, this paper proposes NCS algorithm for squint-looking bistatic SAR, and carefully analysis two key problems in this algorithm: azimuth ambiguity of Doppler center and imaging position in range direction. Finally, it is proved to be valid by the simulation results.
To the big squint mode of bistatic SAR system，based on the echo model of bistatic SAR and range equation in RD area, properly to with the NCS algorithm of conventional SAR, this paper proposes NCS algorithm for squint-looking bistatic SAR, and carefully analysis two key problems in this algorithm: azimuth ambiguity of Doppler center and imaging position in range direction. Finally, it is proved to be valid by the simulation results.
2007, 29(9): 2101-2104.
doi: 10.3724/SP.J.1146.2006.00290
Abstract:
The wide swath synthetic aperture radar (SAR) method basing on elevation multi-receiver is proved to be a high precise and feasible wide swath SAR method by theory and simulation. The SNR of this method has been analyzed by the matrix condition, and the analysis leads to only a qualitative conclusion in previous study. In this paper, the explicit formula is deduced for the SNR of this method, then the limit of the SNR is given and the optimal SNR design rule is presented. Finally, some system schemes according to the optimal SNR rule is given at last and those SNR curve is calculated to validate the rule.
The wide swath synthetic aperture radar (SAR) method basing on elevation multi-receiver is proved to be a high precise and feasible wide swath SAR method by theory and simulation. The SNR of this method has been analyzed by the matrix condition, and the analysis leads to only a qualitative conclusion in previous study. In this paper, the explicit formula is deduced for the SNR of this method, then the limit of the SNR is given and the optimal SNR design rule is presented. Finally, some system schemes according to the optimal SNR rule is given at last and those SNR curve is calculated to validate the rule.
2007, 29(9): 2105-2109.
doi: 10.3724/SP.J.1146.2006.00905
Abstract:
ROI extraction is the essential technology for target recognition in SAR image. This paper firstly analyses the difficulties of ROI extraction for vehicle in SAR image, presents its key steps and flow chart. And then an adaptive CFAR method for SAR image target detection based on the ordered data variability index (ODVI) is proposed to improve the performance of detection. The discrimination features are calculated to form the statistic of target discrimination. The ROIs can be discriminated by using orderly the features. Finally, the real X band SAR image data is used as test data for this ROI extraction algorithm, and the discriminating results are shown. The performance of this algorithm is good.
ROI extraction is the essential technology for target recognition in SAR image. This paper firstly analyses the difficulties of ROI extraction for vehicle in SAR image, presents its key steps and flow chart. And then an adaptive CFAR method for SAR image target detection based on the ordered data variability index (ODVI) is proposed to improve the performance of detection. The discrimination features are calculated to form the statistic of target discrimination. The ROIs can be discriminated by using orderly the features. Finally, the real X band SAR image data is used as test data for this ROI extraction algorithm, and the discriminating results are shown. The performance of this algorithm is good.
2007, 29(9): 2110-2113.
doi: 10.3724/SP.J.1146.2006.00140
Abstract:
The most important specifications of space-borne Synthetic Aperture Radar (SAR) are resolution and swath width. But in conventional SAR, there is a well known trade-off between them because of the pulse mode. In this paper, a new method of multiple azimuth beams is proposed using multiple transmit-receive apertures to further improve the swath width and resolution in azimuth. In addition, it should be noticed that the radiation gain or the equivalent isotropic radiated power per unit area can be greatly increased and the Signal-Noise Rate (SNR) also becomes better. The implement of this method is introduced and comparison with STMAB method and conventional SAR is made with computer simulations, and the effectiveness of this method is illustrated.
The most important specifications of space-borne Synthetic Aperture Radar (SAR) are resolution and swath width. But in conventional SAR, there is a well known trade-off between them because of the pulse mode. In this paper, a new method of multiple azimuth beams is proposed using multiple transmit-receive apertures to further improve the swath width and resolution in azimuth. In addition, it should be noticed that the radiation gain or the equivalent isotropic radiated power per unit area can be greatly increased and the Signal-Noise Rate (SNR) also becomes better. The implement of this method is introduced and comparison with STMAB method and conventional SAR is made with computer simulations, and the effectiveness of this method is illustrated.
2007, 29(9): 2114-2116.
doi: 10.3724/SP.J.1146.2006.00297
Abstract:
This paper studies a correction method for saturated Synthetic Aperture Radar (SAR) raw data based on optimal quantization theory. The method counts the standard deviation of the data and adjusts the data using a certain gain function. The results of simulation indicate this method can do good to radar signal processing.
This paper studies a correction method for saturated Synthetic Aperture Radar (SAR) raw data based on optimal quantization theory. The method counts the standard deviation of the data and adjusts the data using a certain gain function. The results of simulation indicate this method can do good to radar signal processing.
2007, 29(9): 2117-2119.
doi: 10.3724/SP.J.1146.2006.00357
Abstract:
In this paper, the combination of the phase coded stepped frequency and the m-sequences is applied to the distributed space-based radar system using the mode of multi-emitter and single-receiver, and the system achieves excellent range resolution and Doppler resolution. The simulation and performance analysis demonstrate the feasibility of this signal model.
In this paper, the combination of the phase coded stepped frequency and the m-sequences is applied to the distributed space-based radar system using the mode of multi-emitter and single-receiver, and the system achieves excellent range resolution and Doppler resolution. The simulation and performance analysis demonstrate the feasibility of this signal model.
2007, 29(9): 2120-2123.
doi: 10.3724/SP.J.1146.2006.00738
Abstract:
EKF and UKF are often used in target tracking, but the required PDF is approximated by a Gaussian, which may be a gross distortion of the true underlying structure and may lead to filter divergence, especially in the situations where the uncertainty of the measurements is large compared to the uncertainty of process model of tracking. Resample introduces the problem of loss of diversity among the particles with particle filer because the uncertainty of process model is small compared to the uncertainty of the measurements. The SMCEKF and SMCUKF algorithms given in this paper ensure the independency of particles by introducing parallel independent EKF and UKF. The required density of the state vector is represented as a set of random samples and its weights, which is updated and propagated recursively on their own estimate. The performance of the filters is greatly superior to the standard EKF and UKF. Analysis and simulation results of the bearing only tracking problem have proved validity of the algorithms.
EKF and UKF are often used in target tracking, but the required PDF is approximated by a Gaussian, which may be a gross distortion of the true underlying structure and may lead to filter divergence, especially in the situations where the uncertainty of the measurements is large compared to the uncertainty of process model of tracking. Resample introduces the problem of loss of diversity among the particles with particle filer because the uncertainty of process model is small compared to the uncertainty of the measurements. The SMCEKF and SMCUKF algorithms given in this paper ensure the independency of particles by introducing parallel independent EKF and UKF. The required density of the state vector is represented as a set of random samples and its weights, which is updated and propagated recursively on their own estimate. The performance of the filters is greatly superior to the standard EKF and UKF. Analysis and simulation results of the bearing only tracking problem have proved validity of the algorithms.
2007, 29(9): 2124-2127.
doi: 10.3724/SP.J.1146.2006.00189
Abstract:
A new algorithm of signal decompositionModified Adaptive Chirplet Decomposition(MACD) is presented to overcome the deficiency of basis functions with linear time-frequency relations. The chirplet atom is extended to quadratic frequency modulated signal, and it can approach the nonlinear time varying components of the original signal. At the same time, a fast algorithm is presented by computing the Cubic Phase (CP) function of the signal, it is shown that the CP function is concentrated along the instantaneous frequency rate, and the peak of the CP function yields the estimates of the coefficients of quadratic and cubic phase, the time center and the amplitude, Then the initial frequency and the time extent can be obtained by the dechirp technique. The scheme of the algorithm is shown in this paper, and the results of simulated data and echolocation signal from brown bat demonstrate the validity of the method proposed.
A new algorithm of signal decompositionModified Adaptive Chirplet Decomposition(MACD) is presented to overcome the deficiency of basis functions with linear time-frequency relations. The chirplet atom is extended to quadratic frequency modulated signal, and it can approach the nonlinear time varying components of the original signal. At the same time, a fast algorithm is presented by computing the Cubic Phase (CP) function of the signal, it is shown that the CP function is concentrated along the instantaneous frequency rate, and the peak of the CP function yields the estimates of the coefficients of quadratic and cubic phase, the time center and the amplitude, Then the initial frequency and the time extent can be obtained by the dechirp technique. The scheme of the algorithm is shown in this paper, and the results of simulated data and echolocation signal from brown bat demonstrate the validity of the method proposed.
2007, 29(9): 2128-2131.
doi: 10.3724/SP.J.1146.2006.00177
Abstract:
The filterbank without good stopband attenuation will cause the narrowband interference power diffusing in more subbands, and the interference suppressing in transform domain may become less effective and more difficult. Accordingly, a interference suppression method is proposed based on the Near Perfect Reconstruction (NPR) cosine-modulated filterbank，which has more freedom degree in design and is more likely to implement the high stopband attenuation. The filterbank with the characteristic of high stopband attenuation can transform the narrowband interference in less subbands, and therefore has more efficiency in suppressing the narrowband interference by the conventional threshold-detecting. Simulating results show that the performance of the transform-domain interference suppression method using the NPR filterbank is better than the one using the perfect reconstruction filterbanks such as the lapped transform filterbank.
The filterbank without good stopband attenuation will cause the narrowband interference power diffusing in more subbands, and the interference suppressing in transform domain may become less effective and more difficult. Accordingly, a interference suppression method is proposed based on the Near Perfect Reconstruction (NPR) cosine-modulated filterbank，which has more freedom degree in design and is more likely to implement the high stopband attenuation. The filterbank with the characteristic of high stopband attenuation can transform the narrowband interference in less subbands, and therefore has more efficiency in suppressing the narrowband interference by the conventional threshold-detecting. Simulating results show that the performance of the transform-domain interference suppression method using the NPR filterbank is better than the one using the perfect reconstruction filterbanks such as the lapped transform filterbank.
2007, 29(9): 2132-2136.
doi: 10.3724/SP.J.1146.2006.00217
Abstract:
In this paper, two fractionally-spaced space frequency equalization schemes for the downlink of broadband single-carrier multi-user MIMO-CDMA system are proposed. In the first scheme, the fractionally-spaced space frequency equalizer is used to combat the Inter-Code Interference (ICI) and Co-Channel Interference (CCI) induced by MIMO frequency selective fading channels. Under the condition that the spreading codes of all active users are known at the receiver, the second scheme combines fractionally-spaced space frequency equalizer with Parallel Interference Cancellation (PIC), which can reduce the Inter-Antenna Interference (IAI) further. Compared with the conventional frequency domain equalization schemes, the proposed fractionally-spaced space frequency equalization ones have better performance with low complexity.
In this paper, two fractionally-spaced space frequency equalization schemes for the downlink of broadband single-carrier multi-user MIMO-CDMA system are proposed. In the first scheme, the fractionally-spaced space frequency equalizer is used to combat the Inter-Code Interference (ICI) and Co-Channel Interference (CCI) induced by MIMO frequency selective fading channels. Under the condition that the spreading codes of all active users are known at the receiver, the second scheme combines fractionally-spaced space frequency equalizer with Parallel Interference Cancellation (PIC), which can reduce the Inter-Antenna Interference (IAI) further. Compared with the conventional frequency domain equalization schemes, the proposed fractionally-spaced space frequency equalization ones have better performance with low complexity.
2007, 29(9): 2137-2141.
doi: 10.3724/SP.J.1146.2006.00195
Abstract:
This paper proposes a new scheme that introduces cyclic delay diversity into Grouped Linear Constellation Precoding(GLCP) in an OFDM-based systems. By using cyclic delay diversity, a virtual delay spreaded channel can be get, and multipath diversity can be picked up by employing the grouped linear constellation precoding. This new scheme enables low decoding and equalization complexity. According to different channel models, different coding schemes can be designed, and at least get the same spatial diversity as space-time coding with equal number of transmit antennas. It is applicable to any number of transmit antennas with no loss in rate, too. Incorporated with STBC, a new STF coding can be constructed. Simulations reveal the performance merits of the new schemes.
This paper proposes a new scheme that introduces cyclic delay diversity into Grouped Linear Constellation Precoding(GLCP) in an OFDM-based systems. By using cyclic delay diversity, a virtual delay spreaded channel can be get, and multipath diversity can be picked up by employing the grouped linear constellation precoding. This new scheme enables low decoding and equalization complexity. According to different channel models, different coding schemes can be designed, and at least get the same spatial diversity as space-time coding with equal number of transmit antennas. It is applicable to any number of transmit antennas with no loss in rate, too. Incorporated with STBC, a new STF coding can be constructed. Simulations reveal the performance merits of the new schemes.
2007, 29(9): 2142-2145.
doi: 10.3724/SP.J.1146.2006.00147
Abstract:
A communication system With Cooperative Communication based on Space-Time Block Coding (WCC-STBC) is proposed in this paper. A closed-form expression of bit error ratio for WCC-STBC system is also derived. The results through computer Monte Carlo simulation show that the diversity gain, which is dependent on the characteristics of the users individual channel and the cooperative channel between them, can be gotten in the WCC-STBC system. When BER is 10-3, and if the cooperative channel between them is 5dB better than their individuals, about 3dB gain could be gotten in WCC-STBC system; and if one users channel is 5dB better than his partners, 5dB gain for the performance of the worse user could be improved while that of the better one is less loss, only about 0.5dB or so.
A communication system With Cooperative Communication based on Space-Time Block Coding (WCC-STBC) is proposed in this paper. A closed-form expression of bit error ratio for WCC-STBC system is also derived. The results through computer Monte Carlo simulation show that the diversity gain, which is dependent on the characteristics of the users individual channel and the cooperative channel between them, can be gotten in the WCC-STBC system. When BER is 10-3, and if the cooperative channel between them is 5dB better than their individuals, about 3dB gain could be gotten in WCC-STBC system; and if one users channel is 5dB better than his partners, 5dB gain for the performance of the worse user could be improved while that of the better one is less loss, only about 0.5dB or so.
2007, 29(9): 2146-2148.
doi: 10.3724/SP.J.1146.2006.00304
Abstract:
Two universal bounds for pure additive quantum error correcting codes are obtained by using of Krawtchouk polynomials, and it is proved that the quantum Singleton bound and the asymptotic quantum Hamming bound are just the special cases of those two universal bounds.
Two universal bounds for pure additive quantum error correcting codes are obtained by using of Krawtchouk polynomials, and it is proved that the quantum Singleton bound and the asymptotic quantum Hamming bound are just the special cases of those two universal bounds.
2007, 29(9): 2149-2153.
doi: 10.3724/SP.J.1146.2006.00065
Abstract:
In Communication systems, code rates are usually adapted to different channel environment, while puncturing is an effective solution to this strategy. Based on Gaussian approximation, the message-passing decoding algorithm for punctured Low-Density Parity-Check (LDPC)codes on Additive White Gaussian Noise (AWGN) channel is analyzed, as well as the recursive function for message mean is derived. According to this function, a novel design approach is proposed for optimizing the puncturing distributions. Simulation results demonstrate that not only the message mean of message-passing decoder increases more quickly for punctured LDPC optimized by the proposed approach, but also the convergence rate of message-passing decoder is faster than that by traditional approaches.
In Communication systems, code rates are usually adapted to different channel environment, while puncturing is an effective solution to this strategy. Based on Gaussian approximation, the message-passing decoding algorithm for punctured Low-Density Parity-Check (LDPC)codes on Additive White Gaussian Noise (AWGN) channel is analyzed, as well as the recursive function for message mean is derived. According to this function, a novel design approach is proposed for optimizing the puncturing distributions. Simulation results demonstrate that not only the message mean of message-passing decoder increases more quickly for punctured LDPC optimized by the proposed approach, but also the convergence rate of message-passing decoder is faster than that by traditional approaches.
2007, 29(9): 2154-2157.
doi: 10.3724/SP.J.1146.2006.00175
Abstract:
To analyze quickly the motion information of videos, a fast optical flow estimation algorithm for compressed domain is proposed. Firstly the spatial partial derivatives of image luminance are estimated by using two AC coefficients, and then the predictive residual errors and motion vectors of blocks are appended to estimate the temporal partial derivatives. In addition, after the detailed analysis of those macro-blocks that had no forward motion estimation, the motion information is given approximatively relative to their forward reference frames. And then, a amendatory partial derivative estimation is given for the inconsecutive image blocks when occlusion and cut had happened. Finally, optical flow estimation is performed based on the least square method and partial derivative estimation. Experiments indicated that this method can more accurately estimate the optical flow than the L-K method in pixel domain and the exist method in compressed domain. Moreover the proposed method can greatly reduce the compute time than the estimation in pixel domain.
To analyze quickly the motion information of videos, a fast optical flow estimation algorithm for compressed domain is proposed. Firstly the spatial partial derivatives of image luminance are estimated by using two AC coefficients, and then the predictive residual errors and motion vectors of blocks are appended to estimate the temporal partial derivatives. In addition, after the detailed analysis of those macro-blocks that had no forward motion estimation, the motion information is given approximatively relative to their forward reference frames. And then, a amendatory partial derivative estimation is given for the inconsecutive image blocks when occlusion and cut had happened. Finally, optical flow estimation is performed based on the least square method and partial derivative estimation. Experiments indicated that this method can more accurately estimate the optical flow than the L-K method in pixel domain and the exist method in compressed domain. Moreover the proposed method can greatly reduce the compute time than the estimation in pixel domain.
2007, 29(9): 2158-2160.
doi: 10.3724/SP.J.1146.2006.00922
Abstract:
In this paper, a novel slotted-ALOHA algorithm based on Code division is presented for solving the RFID anti-collision problems. Its advantage to the system throughput is analyzed based on its mathematical model and the condition for systems stability is inferred as well. The theoretical and simulated results indicate that the stable scope of SC-ALOHA is broader than that of S-ALOHA and when N is the degree of spread spectrum codes used, the maximum throughput of SC-ALOHA might achieve N times that of S-ALOHA.
In this paper, a novel slotted-ALOHA algorithm based on Code division is presented for solving the RFID anti-collision problems. Its advantage to the system throughput is analyzed based on its mathematical model and the condition for systems stability is inferred as well. The theoretical and simulated results indicate that the stable scope of SC-ALOHA is broader than that of S-ALOHA and when N is the degree of spread spectrum codes used, the maximum throughput of SC-ALOHA might achieve N times that of S-ALOHA.
2007, 29(9): 2161-2167.
doi: 10.3724/SP.J.1146.2005.01606
Abstract:
The noncoherent receiver is attractive for UWB system implementation in lower data rate applications due to its simplicity for implementation, which, however, is accompanied with the performance degradation. There are some factors that affect the performance of noncoherent receiver, among which is the selection of the pre-filter passband width. In this paper, the inter-path interference incurred by the pre-filter is analyzed numerically and statistically, and the optimal passband widths are analyzed as well based on the derivation of closed-form BER performance expression. It is shown that, the impact of IPI due to the pre-filter is insignificant to noncoherent receiver, and there exist different optimal pre-filter passband widths for different channels and input SNRs. However, the #61485;－3dB/－10dB bandwidth of the Gaussian pulse, as well as 2 times of the pulse width reciprocal, could be treated as a suboptimal filter bandwidth, which would only result in a lightly performance degradation.
The noncoherent receiver is attractive for UWB system implementation in lower data rate applications due to its simplicity for implementation, which, however, is accompanied with the performance degradation. There are some factors that affect the performance of noncoherent receiver, among which is the selection of the pre-filter passband width. In this paper, the inter-path interference incurred by the pre-filter is analyzed numerically and statistically, and the optimal passband widths are analyzed as well based on the derivation of closed-form BER performance expression. It is shown that, the impact of IPI due to the pre-filter is insignificant to noncoherent receiver, and there exist different optimal pre-filter passband widths for different channels and input SNRs. However, the #61485;－3dB/－10dB bandwidth of the Gaussian pulse, as well as 2 times of the pulse width reciprocal, could be treated as a suboptimal filter bandwidth, which would only result in a lightly performance degradation.
2007, 29(9): 2168-2172.
doi: 10.3724/SP.J.1146.2005.01607
Abstract:
The non-coherent receiver is attractive for UWB system due to its implementation simplicity, which, however, is also accompanied with its performance degradation. As an effective optimization method, the integration interval selection is usually performed under the single-interval constraint among existed literatures, to the best knowledge of the authors, which not only results in large searching computation complexity but also could not reach the so-called optimal performance. In this paper, a multiple sub-intervals optimization method is proposed based on the derivation of closed-form BER performance expression of the non-coherent receiver and its corresponding modification under small time-bandwidth product condition. It is shown from the numerical results, that the BER performance of the multiple sub-intervals optimization method is obviously superior to the single-interval method, especially in those sparse multipath channels.
The non-coherent receiver is attractive for UWB system due to its implementation simplicity, which, however, is also accompanied with its performance degradation. As an effective optimization method, the integration interval selection is usually performed under the single-interval constraint among existed literatures, to the best knowledge of the authors, which not only results in large searching computation complexity but also could not reach the so-called optimal performance. In this paper, a multiple sub-intervals optimization method is proposed based on the derivation of closed-form BER performance expression of the non-coherent receiver and its corresponding modification under small time-bandwidth product condition. It is shown from the numerical results, that the BER performance of the multiple sub-intervals optimization method is obviously superior to the single-interval method, especially in those sparse multipath channels.
2007, 29(9): 2173-2176.
doi: 10.3724/SP.J.1146.2006.00128
Abstract:
Multiple antenna wireless systems can provide larger channel capacity and better ability to combat fading. The capacity can be further increased by feeding back some channel state information to the transmitter and implementing antenna subset selection. A fast and dynamic transmit antenna subset selection algorithm is proposed in this paper. Instead of the optimal selection algorithm, which needs to compute the capacity of all the antenna subset combinations, the new algorithm has almost the same capacity with lower complexity. It also has larger capacity than the existent static algorithms. Based on bit interleaved code modulation (BICM) technology, space time adaptive bit interleaved code modulation (ST-ABICM) scheme is constructed by adapting data rate per transmit antenna. Simulation results also prove the excellent performance of the scheme.
Multiple antenna wireless systems can provide larger channel capacity and better ability to combat fading. The capacity can be further increased by feeding back some channel state information to the transmitter and implementing antenna subset selection. A fast and dynamic transmit antenna subset selection algorithm is proposed in this paper. Instead of the optimal selection algorithm, which needs to compute the capacity of all the antenna subset combinations, the new algorithm has almost the same capacity with lower complexity. It also has larger capacity than the existent static algorithms. Based on bit interleaved code modulation (BICM) technology, space time adaptive bit interleaved code modulation (ST-ABICM) scheme is constructed by adapting data rate per transmit antenna. Simulation results also prove the excellent performance of the scheme.
2007, 29(9): 2177-2181.
doi: 10.3724/SP.J.1146.2006.00150
Abstract:
A new postfixing scheme for OFDM is proposed, where PP (Pilot Postfix) is inserted in the guard interval after OFDM symbol, and PP-OFDM system is constructed with BICM-ID(Bit-Interleaved Coded Modulation with Iterative Decoding). PP is generated by IFFT to frequency domain pilot symbols in OFDM symbol, and can be combined with those pilots in OFDM symbol at receiver. Channel estimation,equalization and BICM-ID algorithm for PP-OFDM are also proposed. Effective SNR degradation caused by imperfect channel estimation is studied, and a method is proposed to optimize the amplitude of PP. With the optimal PP amplitude, simulations show that PP-OFDM has better channel estimation performance and lower PER (Packet Error Rate) than CP/ZP-OFDM (Cyclic Prefix/Zero Postfix -OFDM).
A new postfixing scheme for OFDM is proposed, where PP (Pilot Postfix) is inserted in the guard interval after OFDM symbol, and PP-OFDM system is constructed with BICM-ID(Bit-Interleaved Coded Modulation with Iterative Decoding). PP is generated by IFFT to frequency domain pilot symbols in OFDM symbol, and can be combined with those pilots in OFDM symbol at receiver. Channel estimation,equalization and BICM-ID algorithm for PP-OFDM are also proposed. Effective SNR degradation caused by imperfect channel estimation is studied, and a method is proposed to optimize the amplitude of PP. With the optimal PP amplitude, simulations show that PP-OFDM has better channel estimation performance and lower PER (Packet Error Rate) than CP/ZP-OFDM (Cyclic Prefix/Zero Postfix -OFDM).
2007, 29(9): 2182-2186.
doi: 10.3724/SP.J.1146.2006.00231
Abstract:
In this paper, the properties of doubly stochastic matrices and doubly stochastic Markov chains are researched from the viewpoints of the classification of states and equilibrium distributions. These properties are applicable to the analysis and modeling of differential frequency hopping G-functions. It is proved that the doubly stochastic matrix model of G-functions is complete in terms of the state transitions uniformity. Based on this mathematic model, a periodic grouping method is proposed to design G-functions by regarding the state transition as a periodic doubly stochastic Markov chain, and it can achieve good performance both in error-correcting and in frequency interval.
In this paper, the properties of doubly stochastic matrices and doubly stochastic Markov chains are researched from the viewpoints of the classification of states and equilibrium distributions. These properties are applicable to the analysis and modeling of differential frequency hopping G-functions. It is proved that the doubly stochastic matrix model of G-functions is complete in terms of the state transitions uniformity. Based on this mathematic model, a periodic grouping method is proposed to design G-functions by regarding the state transition as a periodic doubly stochastic Markov chain, and it can achieve good performance both in error-correcting and in frequency interval.
2007, 29(9): 2187-2190.
doi: 10.3724/SP.J.1146.2006.00223
Abstract:
Based on the discrete-time signal processing, the representation of Doppler effect on code acquisition performance is analyzed and the principle of combining partial-matched filter with FFT is given in this paper. A fast code acquisition architecture by using partial-matched filter with FFT for the large Doppler offset and long PN code spread spectrum system is designed and realized with FPGA and DSP. This architecture provides a very good trade-off between acquisition performance and complexity. Application results show this architecture is very attractive for mobile satellite communication and DS/SS anti-jamming communication.
Based on the discrete-time signal processing, the representation of Doppler effect on code acquisition performance is analyzed and the principle of combining partial-matched filter with FFT is given in this paper. A fast code acquisition architecture by using partial-matched filter with FFT for the large Doppler offset and long PN code spread spectrum system is designed and realized with FPGA and DSP. This architecture provides a very good trade-off between acquisition performance and complexity. Application results show this architecture is very attractive for mobile satellite communication and DS/SS anti-jamming communication.
2007, 29(9): 2191-2194.
doi: 10.3724/SP.J.1146.2006.00172
Abstract:
Fast Frequency Hopping (FFH) communications is an important technology of anti-jamming in military applications. In this paper, anti-jamming performance and implementation of Ultra-Fast Frequency Hopping (UFFH) test-bed are described. The test-bed capable of transmitting data at 96kbps with hopping rates of up to 100khops/s operates in the VHF/UHF band and keeps capability of communication under the condition in which the 60% of system working spectrum is stained. Based on Software Defined Radio (SDR) technology, the architecture uses a Direct Digital frequency Synthesizer (DDS) to implement demodulation and modulation of Frequency Hopping (FH) signal. Furthermore, different arithmetic of digital signal processing and parameters could be software configured because of the high performance Analog-to-Digital Converter (ADC) and the super-heterodyne architecture. Finally, it is showed by the result of measure and simulation that UFFH test-bed has a robust anti-jamming performance.
Fast Frequency Hopping (FFH) communications is an important technology of anti-jamming in military applications. In this paper, anti-jamming performance and implementation of Ultra-Fast Frequency Hopping (UFFH) test-bed are described. The test-bed capable of transmitting data at 96kbps with hopping rates of up to 100khops/s operates in the VHF/UHF band and keeps capability of communication under the condition in which the 60% of system working spectrum is stained. Based on Software Defined Radio (SDR) technology, the architecture uses a Direct Digital frequency Synthesizer (DDS) to implement demodulation and modulation of Frequency Hopping (FH) signal. Furthermore, different arithmetic of digital signal processing and parameters could be software configured because of the high performance Analog-to-Digital Converter (ADC) and the super-heterodyne architecture. Finally, it is showed by the result of measure and simulation that UFFH test-bed has a robust anti-jamming performance.
2007, 29(9): 2195-2198.
doi: 10.3724/SP.J.1146.2006.00039
Abstract:
A blind space-time Multi-User Detection (MUD) algorithm is proposed for multi-antenna Code Division Multiple Access (CDMA) systems. The proposed algorithm exploits the signal structures in both time domain and spatial domain for interference rejection. Specifically, the proposed algorithm integrates Space-Time Coding (STC) with channel estimation by using the structural features of both Alamouti STC and the Multiple Input Multiple Output (MIMO) channel. The Multiple Access Interference (MAI) is suppressed by projecting the received signal onto the orthogonal complement of the estimated interference subspace. Furthermore, the proposed algorithm can adaptively trace the variation of the wireless channels with low complexity. Simulation results show that the algorithm can effectively suppress the MAI of space-time CDMA systems and significantly improve the system performance.
A blind space-time Multi-User Detection (MUD) algorithm is proposed for multi-antenna Code Division Multiple Access (CDMA) systems. The proposed algorithm exploits the signal structures in both time domain and spatial domain for interference rejection. Specifically, the proposed algorithm integrates Space-Time Coding (STC) with channel estimation by using the structural features of both Alamouti STC and the Multiple Input Multiple Output (MIMO) channel. The Multiple Access Interference (MAI) is suppressed by projecting the received signal onto the orthogonal complement of the estimated interference subspace. Furthermore, the proposed algorithm can adaptively trace the variation of the wireless channels with low complexity. Simulation results show that the algorithm can effectively suppress the MAI of space-time CDMA systems and significantly improve the system performance.
2007, 29(9): 2199-2202.
doi: 10.3724/SP.J.1146.2006.01053
Abstract:
In this paper, an improved multiple-symbol differential sphere decoding for differential unitary space-time modulation is presented. According to this algorithm, the sphere decoding process only search the trial symbols with less Maximum Likelihood (ML) metrics rather than all the symbols in the constellation, and an element-by-element calculation method is also presented to compute the ML metric which facilitates an early detection of trial symbol with large ML metric. Simulations show that the presented algorithm can significantly reduce the computational complexity with nearly no performance degradation in moderate SNR range.
In this paper, an improved multiple-symbol differential sphere decoding for differential unitary space-time modulation is presented. According to this algorithm, the sphere decoding process only search the trial symbols with less Maximum Likelihood (ML) metrics rather than all the symbols in the constellation, and an element-by-element calculation method is also presented to compute the ML metric which facilitates an early detection of trial symbol with large ML metric. Simulations show that the presented algorithm can significantly reduce the computational complexity with nearly no performance degradation in moderate SNR range.
2007, 29(9): 2203-2207.
doi: 10.3724/SP.J.1146.2006.00157
Abstract:
In this paper, adaptive modulation and power allocation techniques are investigated for MIMO-OFDM systems. A computationally efficient optimal joint bit and power allocation algorithm is proposed. For implementation simplicity, a simplified sub-optimal algorithm exploiting equal power allocation is presented. Simulation results show that the proposed joint allocation algorithm obtains optimal performance with low computational complexity. Moreover, compared with the optimal algorithm using equal power allocation, the presented sub-optimal algorithm reduces the complexity significantly with slight performance degradation.
In this paper, adaptive modulation and power allocation techniques are investigated for MIMO-OFDM systems. A computationally efficient optimal joint bit and power allocation algorithm is proposed. For implementation simplicity, a simplified sub-optimal algorithm exploiting equal power allocation is presented. Simulation results show that the proposed joint allocation algorithm obtains optimal performance with low computational complexity. Moreover, compared with the optimal algorithm using equal power allocation, the presented sub-optimal algorithm reduces the complexity significantly with slight performance degradation.
2007, 29(9): 2208-2211.
doi: 10.3724/SP.J.1146.2006.00285
Abstract:
GLST is usually used in the downlink of multi-antenna wireless communication systems to increase receiving diversity. In this paper, the transmission orthogonality is analyzed on the base of the system model. Considering that the complexity of the optimal detector is prohibitive, when a higher-order modulation is used, two algorithms are proposed based on sphere decoder to solve the problem. Simulation results show that both the proposed algorithms can approach the performance of the optimal detector.
GLST is usually used in the downlink of multi-antenna wireless communication systems to increase receiving diversity. In this paper, the transmission orthogonality is analyzed on the base of the system model. Considering that the complexity of the optimal detector is prohibitive, when a higher-order modulation is used, two algorithms are proposed based on sphere decoder to solve the problem. Simulation results show that both the proposed algorithms can approach the performance of the optimal detector.
2007, 29(9): 2212-2216.
doi: 10.3724/SP.J.1146.2006.00183
Abstract:
Based on the positive bias property of the Time-Of-Arrival(TOA)measurement error caused by Non- Line-Of-Sight(NLOS)propagation in the mobile communication environment, a simple geometrical position location method is presented. Some sites are determined according to the coordinates of the base-stations and TOAs, and the location estimation of mobile station is achieved by averaging the coordinates of these sites. The simulation results show that this method can effectively improve the location accuracy in NLOS environment with little degeneration to mobile location in LOS environment.
Based on the positive bias property of the Time-Of-Arrival(TOA)measurement error caused by Non- Line-Of-Sight(NLOS)propagation in the mobile communication environment, a simple geometrical position location method is presented. Some sites are determined according to the coordinates of the base-stations and TOAs, and the location estimation of mobile station is achieved by averaging the coordinates of these sites. The simulation results show that this method can effectively improve the location accuracy in NLOS environment with little degeneration to mobile location in LOS environment.
2007, 29(9): 2217-2221.
doi: 10.3724/SP.J.1146.2006.00143
Abstract:
Application layer mobility is important for IP application under mobile environment. The support of under layer mobility guarantees the smooth of data transaction in handoff, while application layer mobility guarantees higher layer application can transfer and receive data correctly in handoff. MESOCK (Mobile Extension SOCKET) is introduced for application layer mobility, which is an extension of TCP/UDP and let running session using new IP address to transfer and receive data after getting it. So it can let data go to the destination not be forwarded by tunnel. It makes the datas integrity and continuity in the IP changing process. The result of simulation proves that the MESOCK can get the goal well.
Application layer mobility is important for IP application under mobile environment. The support of under layer mobility guarantees the smooth of data transaction in handoff, while application layer mobility guarantees higher layer application can transfer and receive data correctly in handoff. MESOCK (Mobile Extension SOCKET) is introduced for application layer mobility, which is an extension of TCP/UDP and let running session using new IP address to transfer and receive data after getting it. So it can let data go to the destination not be forwarded by tunnel. It makes the datas integrity and continuity in the IP changing process. The result of simulation proves that the MESOCK can get the goal well.
2007, 29(9): 2222-2225.
doi: 10.3724/SP.J.1146.2006.01568
Abstract:
The linear complexity and the k-errorlinear complexity of a periodic sequence are two important standards to scale the randomicity of keystreams. For a pn-periodic sequence over Fp, the possible values of the 1-error linear complexity and the number of sequences with certain 1-error linear complexity are established. Moreover, the expected value of the 1-error linear complexity for a random pn-periodic sequence over Fp is also given.
The linear complexity and the k-errorlinear complexity of a periodic sequence are two important standards to scale the randomicity of keystreams. For a pn-periodic sequence over Fp, the possible values of the 1-error linear complexity and the number of sequences with certain 1-error linear complexity are established. Moreover, the expected value of the 1-error linear complexity for a random pn-periodic sequence over Fp is also given.
2007, 29(9): 2226-2229.
doi: 10.3724/SP.J.1146.2006.00288
Abstract:
A steganographic method is proposed for embedding secret information in digital speech signals. Depending on the PCM A/ law compression/decompression characteristics and mutual conversion between the two laws, appropriate bit planes are chosen in modifying the host speech samples to embed secret data. The induced distortion is highly imperceptible, and the embedded information is robust against A/ law compression and mutual conversion, therefore suitable for transmission not only through the Internet but also through international PSTN. Experimental results indicate the effectiveness of the proposed method.
A steganographic method is proposed for embedding secret information in digital speech signals. Depending on the PCM A/ law compression/decompression characteristics and mutual conversion between the two laws, appropriate bit planes are chosen in modifying the host speech samples to embed secret data. The induced distortion is highly imperceptible, and the embedded information is robust against A/ law compression and mutual conversion, therefore suitable for transmission not only through the Internet but also through international PSTN. Experimental results indicate the effectiveness of the proposed method.
2007, 29(9): 2230-2234.
doi: 10.3724/SP.J.1146.2006.01163
Abstract:
In this paper, the Transient Chaotic Neural Network(TCNN) is used to solve the Channel Assignment Problem(CAP), and a new method named two-stage annealing method in TCNN is proposed. The neural network gradually convergences, through the transient chaos, to a stable equilibrium point according to the damping of the self-feedback connection weight, and the dividing point in the new model is chosen according to the change of the corresponding Lyapunov exponent. The two- stage annealing method can make sure the network take good advantage of the chaos to search the global minimum and enhance the convergence rate. In the 7-cell cellular network, the convergence rate is 30% higher than the TCNN model，and is also upgraded 15% in the Kunzs benchmark test. Simulated results show that the new model has a higher searching ability and lower computing time in searching the global minimum. The searching ability and the choosing of the parameters are also discussed based on the simulated results.
In this paper, the Transient Chaotic Neural Network(TCNN) is used to solve the Channel Assignment Problem(CAP), and a new method named two-stage annealing method in TCNN is proposed. The neural network gradually convergences, through the transient chaos, to a stable equilibrium point according to the damping of the self-feedback connection weight, and the dividing point in the new model is chosen according to the change of the corresponding Lyapunov exponent. The two- stage annealing method can make sure the network take good advantage of the chaos to search the global minimum and enhance the convergence rate. In the 7-cell cellular network, the convergence rate is 30% higher than the TCNN model，and is also upgraded 15% in the Kunzs benchmark test. Simulated results show that the new model has a higher searching ability and lower computing time in searching the global minimum. The searching ability and the choosing of the parameters are also discussed based on the simulated results.
2007, 29(9): 2235-2238.
doi: 10.3724/SP.J.1146.2006.00294
Abstract:
As continuing growth in traffic volume from users with varying degrees of mobility in wireless networks, a Hierarchical Cellular Structures (HCS) has been suggested. In this paper, the performance of a two-tier microcell/macrocel1 cellular system with bidirectional call-overflow scheme is studied, where cell selection in hierarchical architecture is based on the velocity of the mobile making the calls. Different from the general layer-classified (macrocell and microcell) analysis, a user-classified (fast and slow) modeling method is developed for evaluating the performance of such systems. The user-classified models consist of a fast user model and a slow user model both of which are simple one-dimension Markov process transition diagrams. Theoretical analysis and numerical evaluation show that the analytical models are reliable. Furthermore, effect of velocity threshold and blocked call repeat phenomenon using the models are examined.
As continuing growth in traffic volume from users with varying degrees of mobility in wireless networks, a Hierarchical Cellular Structures (HCS) has been suggested. In this paper, the performance of a two-tier microcell/macrocel1 cellular system with bidirectional call-overflow scheme is studied, where cell selection in hierarchical architecture is based on the velocity of the mobile making the calls. Different from the general layer-classified (macrocell and microcell) analysis, a user-classified (fast and slow) modeling method is developed for evaluating the performance of such systems. The user-classified models consist of a fast user model and a slow user model both of which are simple one-dimension Markov process transition diagrams. Theoretical analysis and numerical evaluation show that the analytical models are reliable. Furthermore, effect of velocity threshold and blocked call repeat phenomenon using the models are examined.
2007, 29(9): 2239-2243.
doi: 10.3724/SP.J.1146.2006.00168
Abstract:
By taking the popularity character of the media program and the cost of the transmission network into consideration, a novel caching utility function is proposed. A Network Cost Based streaming media cache allocation and replacement (NCB) algorithm is designed for multi-server streaming application. Simulation proves that, by employing the NCB algorithm, the cache hit rate can be improved obviously and the total network cost can be decreased effectively. In the Internet streaming application environment, where the network connections between content servers and caching proxy are heterogeneous, the amount of streaming programs is huge, the NCB shows good performance.
By taking the popularity character of the media program and the cost of the transmission network into consideration, a novel caching utility function is proposed. A Network Cost Based streaming media cache allocation and replacement (NCB) algorithm is designed for multi-server streaming application. Simulation proves that, by employing the NCB algorithm, the cache hit rate can be improved obviously and the total network cost can be decreased effectively. In the Internet streaming application environment, where the network connections between content servers and caching proxy are heterogeneous, the amount of streaming programs is huge, the NCB shows good performance.
2007, 29(9): 2244-2247.
doi: 10.3724/SP.J.1146.2006.01535
Abstract:
A novel adaptive data fusion routing algorithm based on genetic algorithm is proposed in order to solve the problem of mobile agent collecting correlated data with energy validity. When selecting the data fusion routing, mobile agent moving to every sensor node that shall perform fusion is adaptively adjusted according to data transmission cost, data fusion cost and energy gain. The simulation results indicate that the energy efficiency of adaptive fusion routing algorithm has better performance than full fusion routing algorithm and local closest first heuristics algorithm.
A novel adaptive data fusion routing algorithm based on genetic algorithm is proposed in order to solve the problem of mobile agent collecting correlated data with energy validity. When selecting the data fusion routing, mobile agent moving to every sensor node that shall perform fusion is adaptively adjusted according to data transmission cost, data fusion cost and energy gain. The simulation results indicate that the energy efficiency of adaptive fusion routing algorithm has better performance than full fusion routing algorithm and local closest first heuristics algorithm.
2007, 29(9): 2252-2256.
doi: 10.3724/SP.J.1146.2006.00153
Abstract:
Current user behavior models in streaming application ignore user interactivity, a new model is proposed for interactive user behavior in steaming-on-demand. Simulation results show that it can generate synthetic traces with different interactivity degree and user request mode. The statistical characteristics of main objects in synthetic traces are more reasonable, and the inner content popularity approximates actual interactive user traces well.
Current user behavior models in streaming application ignore user interactivity, a new model is proposed for interactive user behavior in steaming-on-demand. Simulation results show that it can generate synthetic traces with different interactivity degree and user request mode. The statistical characteristics of main objects in synthetic traces are more reasonable, and the inner content popularity approximates actual interactive user traces well.
2007, 29(9): 2257-2261.
doi: 10.3724/SP.J.1146.2006.00173
Abstract:
A theoretic analysis is proposed which indirectly proves how vanishing moments influence on estimation Hurst parameter of self-similar traffic, and a lots of experimental results show the Hursts value changes periodicity with the vanishing moments, but total way is the Hursts value decrease and its variance raises with the increase of the vanishing moments. This conclusion is different from common view, and could provide reason for choice vanishing moments on wavelet analysis method for Hurst parameter.
A theoretic analysis is proposed which indirectly proves how vanishing moments influence on estimation Hurst parameter of self-similar traffic, and a lots of experimental results show the Hursts value changes periodicity with the vanishing moments, but total way is the Hursts value decrease and its variance raises with the increase of the vanishing moments. This conclusion is different from common view, and could provide reason for choice vanishing moments on wavelet analysis method for Hurst parameter.
2007, 29(9): 2262-2264.
doi: 10.3724/SP.J.1146.2006.00284
Abstract:
When nodes are organized as clusters, data packets can be transmitted from cluster members to head in either single hop or multi-hop communication mode. Using maximum energy dissipation of cluster members as evaluation criterion, energy efficiency of these two modes is researched. The conclusion was come to that the optimal communication mode depends on channel path loss. Taking the uneven energy consumption caused by solely single hop or multi-hop mode into consideration, a hybrid communication mode that nodes alternate between single hop and multi-hop mode with certain probability is proposed. The simulation result verifies that hybrid mode provides the best energy efficiency as well as even energy dissipation.
When nodes are organized as clusters, data packets can be transmitted from cluster members to head in either single hop or multi-hop communication mode. Using maximum energy dissipation of cluster members as evaluation criterion, energy efficiency of these two modes is researched. The conclusion was come to that the optimal communication mode depends on channel path loss. Taking the uneven energy consumption caused by solely single hop or multi-hop mode into consideration, a hybrid communication mode that nodes alternate between single hop and multi-hop mode with certain probability is proposed. The simulation result verifies that hybrid mode provides the best energy efficiency as well as even energy dissipation.
2007, 29(9): 2265-2268.
doi: 10.3724/SP.J.1146.2006.00628
Abstract:
The Higher-Order Method of Moments (HO-MoM) is frequently used to calculate the electromagnetic scattering, because of its high efficient and accurate. Several issues are analyzed in this paper for using the HO-MoM. Firstly, multi-dimension numerical integrations calculated again and again, a novel method is presented in this paper to translate integrations to the product of matrices. When the matrix from MoM filled, it is only needed to calculated the Greens function at integral nodes. That will reduces CPU time and increase the efficiency of HO-MoM. Secondly, the different methods, used to treat with singular integral or quasi-singular integral, is summarized and estimated.
The Higher-Order Method of Moments (HO-MoM) is frequently used to calculate the electromagnetic scattering, because of its high efficient and accurate. Several issues are analyzed in this paper for using the HO-MoM. Firstly, multi-dimension numerical integrations calculated again and again, a novel method is presented in this paper to translate integrations to the product of matrices. When the matrix from MoM filled, it is only needed to calculated the Greens function at integral nodes. That will reduces CPU time and increase the efficiency of HO-MoM. Secondly, the different methods, used to treat with singular integral or quasi-singular integral, is summarized and estimated.
2007, 29(9): 2269-2273.
doi: 10.3724/SP.J.1146.2006.00101
Abstract:
The method combined IPO, FMM, RPFMM, FaFFA can rapidly solve the scattering of three-dimensional cavity. In this paper, the Generalized Reciprocity Integral(GRI) method is applied to treat the cavities with termination. The cavity is divide into two different parts by a surface St which is located sufficiently close to the termination, the front section (typically simple and smooth) is analyzed by a hybrid method which IPO combined with FMM, RPFMM and FaFFA, the termination can be analyzed independently. To solve the deep cavities and improve computing speed the front section of cavity is subdivided into several sections, and each of them is analyzed independently using the IPO combined fast computing method from the rest of the cavity. Then a connection scheme is used to get the scattering field radiated by the front section at the surface St . Finally the RCS of the cavity is get by the GRI at surface St. The numerical results show the hybrid method is accurate enough and can improve effectively computing speed。
The method combined IPO, FMM, RPFMM, FaFFA can rapidly solve the scattering of three-dimensional cavity. In this paper, the Generalized Reciprocity Integral(GRI) method is applied to treat the cavities with termination. The cavity is divide into two different parts by a surface St which is located sufficiently close to the termination, the front section (typically simple and smooth) is analyzed by a hybrid method which IPO combined with FMM, RPFMM and FaFFA, the termination can be analyzed independently. To solve the deep cavities and improve computing speed the front section of cavity is subdivided into several sections, and each of them is analyzed independently using the IPO combined fast computing method from the rest of the cavity. Then a connection scheme is used to get the scattering field radiated by the front section at the surface St . Finally the RCS of the cavity is get by the GRI at surface St. The numerical results show the hybrid method is accurate enough and can improve effectively computing speed。
2007, 29(9): 2274-2278.
doi: 10.3724/SP.J.1146.2006.00043
Abstract:
In the modeling of CoPlanar Waveguide (CPW) by a quasi-static Finite Difference Method (FDM), to resolve the problem of too much simulation time with equidistant discretization in the 3D FDM simulation and improve the non-equidistant discretization in the algorithm complexity and efficiency, a new progressive discretization is put forward, which is extremely easy to implement with good compromise of precision and efficiency. And the CPW transmission line, open end and short end are modeled successfully based on the improved FDM. After a special modeling of CPW short end, the progressive discretization is proved to be practicable. At the same time, some particular references are also presented as in some applications.
In the modeling of CoPlanar Waveguide (CPW) by a quasi-static Finite Difference Method (FDM), to resolve the problem of too much simulation time with equidistant discretization in the 3D FDM simulation and improve the non-equidistant discretization in the algorithm complexity and efficiency, a new progressive discretization is put forward, which is extremely easy to implement with good compromise of precision and efficiency. And the CPW transmission line, open end and short end are modeled successfully based on the improved FDM. After a special modeling of CPW short end, the progressive discretization is proved to be practicable. At the same time, some particular references are also presented as in some applications.
2007, 29(9): 2279-2282.
doi: 10.3724/SP.J.1146.2006.00346
Abstract:
Multi-capture composition is a valid method on extending the dynamic range of CMOS image sensor (CIS). In this paper, multi-capture composition using in real-time control is studied. By acquired the average value of all pixels, the number of saturated pixels and severe underexposure pixels, the control mode， the number of exposure， integrating time of each exposure and composition algorithm is determined in real-time. This method has been implemented in a 640480 CIS and simulated with FPGA. The simulated result indicates this method can extend the dynamic range effectively.
Multi-capture composition is a valid method on extending the dynamic range of CMOS image sensor (CIS). In this paper, multi-capture composition using in real-time control is studied. By acquired the average value of all pixels, the number of saturated pixels and severe underexposure pixels, the control mode， the number of exposure， integrating time of each exposure and composition algorithm is determined in real-time. This method has been implemented in a 640480 CIS and simulated with FPGA. The simulated result indicates this method can extend the dynamic range effectively.
2007, 29(9): 2283-2286.
doi: 10.3724/SP.J.1146.2006.01773
Abstract:
The crosstalk of T-shaped microstrip is the important component in signal integrity of a high-speed circuit. In this paper, the fullwave analysis of the crosstalk is carried out by adopting the Perfectly Matched Layer Finite-Difference Time-Domain (PML-FDTD) method, and then the measures of reducing crosstalk are proposed in detail. The different structure parameters of improved T-shaped microstrip have different influence on the crosstalk. That is, (1) the crosstalk between microstrips is smaller as the orifice distance S is becoming bigger, and the amplitude of crosstalk is reduced to 10dB; (2) The value of is smaller as the crosstalk between microstrip is becoming smaller correspondingly, and the amplitude of crosstalk is reduced to 14dB. Thus opening an orifice on the T-shaped microstrip can reduce the crosstalk between one T-shaped microstrip and another adjacent one remarkably by changing orifice distance S and angle .
The crosstalk of T-shaped microstrip is the important component in signal integrity of a high-speed circuit. In this paper, the fullwave analysis of the crosstalk is carried out by adopting the Perfectly Matched Layer Finite-Difference Time-Domain (PML-FDTD) method, and then the measures of reducing crosstalk are proposed in detail. The different structure parameters of improved T-shaped microstrip have different influence on the crosstalk. That is, (1) the crosstalk between microstrips is smaller as the orifice distance S is becoming bigger, and the amplitude of crosstalk is reduced to 10dB; (2) The value of is smaller as the crosstalk between microstrip is becoming smaller correspondingly, and the amplitude of crosstalk is reduced to 14dB. Thus opening an orifice on the T-shaped microstrip can reduce the crosstalk between one T-shaped microstrip and another adjacent one remarkably by changing orifice distance S and angle .
2007, 29(9): 2248-2251.
doi: 10.3724/SP.J.1146.2005.01632
Abstract:
At present some Intrusion Detection Systems (IDS) use the features of TCP/IP data packets for analysis and modeling, but due to the different contribution of TCP/IP features to the detecting process a favorable impact may be made on the promotion of IDSs detecting rate and real time if the quantity of properties can be reduced properly without affecting the precision of detection. Therefore, a Decision Tree Rule-based Statistical method (DTRS) in light of this is presented to reduce TCP/IP features. Its primary concept is to create n decision trees in n data subsets, extract the rules, work out the relatively important features in accordance with the frequency of use of different features and verify its feasibility and effectiveness through tests.
At present some Intrusion Detection Systems (IDS) use the features of TCP/IP data packets for analysis and modeling, but due to the different contribution of TCP/IP features to the detecting process a favorable impact may be made on the promotion of IDSs detecting rate and real time if the quantity of properties can be reduced properly without affecting the precision of detection. Therefore, a Decision Tree Rule-based Statistical method (DTRS) in light of this is presented to reduce TCP/IP features. Its primary concept is to create n decision trees in n data subsets, extract the rules, work out the relatively important features in accordance with the frequency of use of different features and verify its feasibility and effectiveness through tests.
