Email alert
2010 Vol. 32, No. 5
Display Method:
2010, 32(5): 1017-1022.
doi: 10.3724/SP.J.1146.2009.00552
Abstract:
A novel target automatic detection algorithm is proposed in this paper, and it is mainly used for the processing of the man-made targets with a relatively complex structure in natural scenes images and high-resolution remote sensing images. Based on each geometric component of objects, this method needs less training samples. First of all, it selects two sorts of typical features and trains classifiers by machine learning correspondingly, which can effectively prevent the decrease of accuracy for the similarities between interest objects and some objects in background. Then the method detects targets top-down and automatically with the marked point process model, whose data terms consist of the priori constraint on the objects distribution and respondences of trained classifiers. The experimental results demonstrate the precision, robustness, and effectiveness of the proposed method.
A novel target automatic detection algorithm is proposed in this paper, and it is mainly used for the processing of the man-made targets with a relatively complex structure in natural scenes images and high-resolution remote sensing images. Based on each geometric component of objects, this method needs less training samples. First of all, it selects two sorts of typical features and trains classifiers by machine learning correspondingly, which can effectively prevent the decrease of accuracy for the similarities between interest objects and some objects in background. Then the method detects targets top-down and automatically with the marked point process model, whose data terms consist of the priori constraint on the objects distribution and respondences of trained classifiers. The experimental results demonstrate the precision, robustness, and effectiveness of the proposed method.
2010, 32(5): 1023-1028.
doi: 10.3724/SP.J.1146.2009.01275
Abstract:
The centerline as a distinct feature in ship ISAR image, its precision affects the recognition result. In this paper, a new ship centerline extraction algorithm is presented. According to this algorithm, the ship width and centerline are together detected by line group, and the parameters are optimized according to the radius and angle sensitivity of ship area in the line group. The simulation results demonstrate that this algorithm has higher precision, can adapt to images with different projection plane.
The centerline as a distinct feature in ship ISAR image, its precision affects the recognition result. In this paper, a new ship centerline extraction algorithm is presented. According to this algorithm, the ship width and centerline are together detected by line group, and the parameters are optimized according to the radius and angle sensitivity of ship area in the line group. The simulation results demonstrate that this algorithm has higher precision, can adapt to images with different projection plane.
2010, 32(5): 1029-1033.
doi: 10.3724/SP.J.1146.2009.00737
Abstract:
In the applications of airborne multi-baseline SAR tomography, the baseline is needed to be small while the aperture is required to be large. However, it is difficult to meet the two conditions simultaneously. In this paper, the sparse signal representation is used to solve the SAR tomography with few baselines. Firstly, the targets are modeled as sparse based on the hypothesis of multi-center of scattering. And then, the redundant dictionary is constructed through the geometry of SAR observing system and the frequencies of signal. As a result, the SAR tomography is transformed into the problem of sparse signal representation. The imaging results are obtained by solving the ill-posed problem based on sparse regularization. The relationship between the parameters and the quality of imaging is shown through the processing of the simulated serial signal. Finally, the validity of algorithm is clarified by the results of the three-dimensional SAR tomography.
In the applications of airborne multi-baseline SAR tomography, the baseline is needed to be small while the aperture is required to be large. However, it is difficult to meet the two conditions simultaneously. In this paper, the sparse signal representation is used to solve the SAR tomography with few baselines. Firstly, the targets are modeled as sparse based on the hypothesis of multi-center of scattering. And then, the redundant dictionary is constructed through the geometry of SAR observing system and the frequencies of signal. As a result, the SAR tomography is transformed into the problem of sparse signal representation. The imaging results are obtained by solving the ill-posed problem based on sparse regularization. The relationship between the parameters and the quality of imaging is shown through the processing of the simulated serial signal. Finally, the validity of algorithm is clarified by the results of the three-dimensional SAR tomography.
2010, 32(5): 1034-1039.
doi: 10.3724/SP.J.1146.2009.00582
Abstract:
The inherent errors of radar system degrade seriously the performance of Synthetic Aperture Radar(SAR), and these errors should be calibrated to acquire high quality images. In this paper, to the issue of the impact of system errors and non-linearities in FMCW SAR, models of system errors are established, and then the estimation and calibration of system errors are investigated in detail. An improved non-linear high resolution range-Doppler algorithm for FMCW SAR is proposed, according to the requirements of mini-architecture and real-time in FMCM SAR imaging systems. Finally, the feasibility and validity of the presented algorithm is demonstrated with simulation.
The inherent errors of radar system degrade seriously the performance of Synthetic Aperture Radar(SAR), and these errors should be calibrated to acquire high quality images. In this paper, to the issue of the impact of system errors and non-linearities in FMCW SAR, models of system errors are established, and then the estimation and calibration of system errors are investigated in detail. An improved non-linear high resolution range-Doppler algorithm for FMCW SAR is proposed, according to the requirements of mini-architecture and real-time in FMCM SAR imaging systems. Finally, the feasibility and validity of the presented algorithm is demonstrated with simulation.
2010, 32(5): 1040-1044.
doi: 10.3724/SP.J.1146.2009.00733
Abstract:
In this paper, an algorithm for compressing SAR raw data is proposed. This algorithm is based on the combination of Discrete Cosine Transform(DCT) and Trellis Coding Quantization(TCQ). Because SAR raw data consists of mass linear frequency modulation signals and have plenty of frequency components, a quantization bit allocation method is desired with grouped DCT coefficients based on frequency components, and quantizing the coefficients with TCQ methods. therefore this algorithm obtains notable quantization gain. Simulation results on real SAR raw data show that the proposed algorithm outperforms existed algorithms.
In this paper, an algorithm for compressing SAR raw data is proposed. This algorithm is based on the combination of Discrete Cosine Transform(DCT) and Trellis Coding Quantization(TCQ). Because SAR raw data consists of mass linear frequency modulation signals and have plenty of frequency components, a quantization bit allocation method is desired with grouped DCT coefficients based on frequency components, and quantizing the coefficients with TCQ methods. therefore this algorithm obtains notable quantization gain. Simulation results on real SAR raw data show that the proposed algorithm outperforms existed algorithms.
2010, 32(5): 1045-1049.
doi: 10.3724/SP.J.1146.2009.01215
Abstract:
Adaptive Pulse Compression (APC) can effectively mitigate the mutual interference between the transmitting waveforms of Multiple-Input Multiple-Output (MIMO) radar. In this paper, APC of MIMO radar based on Multi-Stage Wiener Filter (MSWF) is proposed, in which, the weights of forward and backward recursion of MSWF are computed to obtain the weight coefficients of the APC. The proposed method does not involve the estimate and the inverse of the sample covariance matrix, thus indicating much lower computation complexity. The numerical results indicate that the proposed method is superior to the conventional matched filter, and performs similarly as the APC based on MMSE and GSC.
Adaptive Pulse Compression (APC) can effectively mitigate the mutual interference between the transmitting waveforms of Multiple-Input Multiple-Output (MIMO) radar. In this paper, APC of MIMO radar based on Multi-Stage Wiener Filter (MSWF) is proposed, in which, the weights of forward and backward recursion of MSWF are computed to obtain the weight coefficients of the APC. The proposed method does not involve the estimate and the inverse of the sample covariance matrix, thus indicating much lower computation complexity. The numerical results indicate that the proposed method is superior to the conventional matched filter, and performs similarly as the APC based on MMSE and GSC.
2010, 32(5): 1050-1055.
doi: 10.3724/SP.J.1146.2009.00669
Abstract:
It is important to classify and recognize ground moving targets such as armored cars. Micro-Doppler features have special advantages in identification due to the detailed information of target. The micro-Doppler features of turret and gun tube which possessed by tank only are discussed, a method or general process for parameter estimation and precise identification based on micro-Doppler features of different tank types is proposed, which can be valuable assistant technical approach to existing methods. Validity is verified with the identicalness of calculated and simulated time-frequency spectrum.
It is important to classify and recognize ground moving targets such as armored cars. Micro-Doppler features have special advantages in identification due to the detailed information of target. The micro-Doppler features of turret and gun tube which possessed by tank only are discussed, a method or general process for parameter estimation and precise identification based on micro-Doppler features of different tank types is proposed, which can be valuable assistant technical approach to existing methods. Validity is verified with the identicalness of calculated and simulated time-frequency spectrum.
2010, 32(5): 1056-1060.
doi: 10.3724/SP.J.1146.2009.00239
Abstract:
With the requirement of high resolution and wide swath, multiple-transit and multiple-receive SAR becomes a important research field, multiple apertures use the same carrier wave frequency, transit orthogonal phase coding signal is one of modes, which can be implemented in the engineering . Through the particular research of the echo array and the echo phase compensate four kinds impossible methods of echo processing and the signal matrixes of the methods are proposed. The simulation with orthogonal chaotic coding signal shows the effectiveness of the methods.
With the requirement of high resolution and wide swath, multiple-transit and multiple-receive SAR becomes a important research field, multiple apertures use the same carrier wave frequency, transit orthogonal phase coding signal is one of modes, which can be implemented in the engineering . Through the particular research of the echo array and the echo phase compensate four kinds impossible methods of echo processing and the signal matrixes of the methods are proposed. The simulation with orthogonal chaotic coding signal shows the effectiveness of the methods.
2010, 32(5): 1061-1065.
doi: 10.3724/SP.J.1146.2009.00577
Abstract:
The traditional image quality assessment methods based on pixels have their own limitations, such as the lack of consideration of the image structure, or the need of a complete reference image. To avoid these problems, this paper presented a new image quality assessment method based on Block Weighted Singular Value Decomposition (BWSVD). First, the images are divided into blocks at size 88, then, the singular value vector difference and the mean bias between the original image blocks and the distorted image blocks are considered to evaluate the distortion degree. Beside this, the Human Visual Sensibility (HVS) is considered to determine the weight of each block. Many tests are conducted to evaluate the performance; the 227 testing images are coming from the Live Image Quality Assessment Database, Release 2005. Compared with the PSNR, RMSE, UQI, MSSIM, MSVD algorithms, the presented method shows a great improvement in both the consistency with the DMOS (Differential Mean Opinion Score, DMOS) and the stability when it is applied to different compression rates.
The traditional image quality assessment methods based on pixels have their own limitations, such as the lack of consideration of the image structure, or the need of a complete reference image. To avoid these problems, this paper presented a new image quality assessment method based on Block Weighted Singular Value Decomposition (BWSVD). First, the images are divided into blocks at size 88, then, the singular value vector difference and the mean bias between the original image blocks and the distorted image blocks are considered to evaluate the distortion degree. Beside this, the Human Visual Sensibility (HVS) is considered to determine the weight of each block. Many tests are conducted to evaluate the performance; the 227 testing images are coming from the Live Image Quality Assessment Database, Release 2005. Compared with the PSNR, RMSE, UQI, MSSIM, MSVD algorithms, the presented method shows a great improvement in both the consistency with the DMOS (Differential Mean Opinion Score, DMOS) and the stability when it is applied to different compression rates.
2010, 32(5): 1066-1070.
doi: 10.3724/SP.J.1146.2009.00750
Abstract:
Based on the uniqueness condition of the solution of Nonorthogonal Joint Diagonalization (NJD), the identifiability for Blind Signal Separation (BSS) is analyzed. Firstly, it is proved that the target matrices consisting of Second-Order Statistics (SOS) or higher-order cumulant are diagonalizable, so the problem of BSS can be solved by NJD. The uniqueness condition for NJD is that the vectors consisting of diagonal elements in the same position of diagonal matrix are pairwise linearly independent. From this proposition，the necessary and sufficient condition for BSS is deduced. For second-order statistics based BSS, the condition is that the source signals have not the identical autocorrelation shape. For higher-order cumulant, there is not Gaussian signal in sources. The above conclusion provides a mathematical foundation for the BSS methods based on the NJD. Numerical simulations confirm the conclusion in this paper.
Based on the uniqueness condition of the solution of Nonorthogonal Joint Diagonalization (NJD), the identifiability for Blind Signal Separation (BSS) is analyzed. Firstly, it is proved that the target matrices consisting of Second-Order Statistics (SOS) or higher-order cumulant are diagonalizable, so the problem of BSS can be solved by NJD. The uniqueness condition for NJD is that the vectors consisting of diagonal elements in the same position of diagonal matrix are pairwise linearly independent. From this proposition，the necessary and sufficient condition for BSS is deduced. For second-order statistics based BSS, the condition is that the source signals have not the identical autocorrelation shape. For higher-order cumulant, there is not Gaussian signal in sources. The above conclusion provides a mathematical foundation for the BSS methods based on the NJD. Numerical simulations confirm the conclusion in this paper.
2010, 32(5): 1071-1076.
doi: 10.3724/SP.J.1146.2009.00513
Abstract:
For a special signal environment, where all of the sources have identical nonzero mean values, taking advantages of the Vandermonde structured steering vector of the uniform linear antenna array, an analytical solution of null-subspace is educed in this paper. Based on the solution, both of one-dimensional signal Direction Of Arrival (DOA) estimation algorithm with uniform linear array and two-dimensional signal DOA estimation algorithm with sparse planar array are proposed. The algorithm does not involve the covariance matrix calculation and any matrix decomposition, so it is much computational saved than the conventional super-resolution DOA estimation algorithms. Beside that, the novel method works well either the signal sources are coherence or not. The simulation shows that the novel algorithm possesses the same performance with Root-MUSIC algorithm as the whole when the mean value of the noise is zero and the number of snapshots is enough, and it performs better than Root-Music when the Signal Noise Ratio (SNR) is lower.
For a special signal environment, where all of the sources have identical nonzero mean values, taking advantages of the Vandermonde structured steering vector of the uniform linear antenna array, an analytical solution of null-subspace is educed in this paper. Based on the solution, both of one-dimensional signal Direction Of Arrival (DOA) estimation algorithm with uniform linear array and two-dimensional signal DOA estimation algorithm with sparse planar array are proposed. The algorithm does not involve the covariance matrix calculation and any matrix decomposition, so it is much computational saved than the conventional super-resolution DOA estimation algorithms. Beside that, the novel method works well either the signal sources are coherence or not. The simulation shows that the novel algorithm possesses the same performance with Root-MUSIC algorithm as the whole when the mean value of the noise is zero and the number of snapshots is enough, and it performs better than Root-Music when the Signal Noise Ratio (SNR) is lower.
2010, 32(5): 1077-1082.
doi: 10.3724/SP.J.1146.2009.00613
Abstract:
In the traditional characteristic waveform interpolation speech coding, the high frequency harmonics of the synthetic speech are usually lost, which makes the speech feel noisy, due to the phase information neglect of the characteristic waveform and the whole characteristic waveform alignment. In order to improve the synthetic speech quality, the multiband surd/sonant flag is first introduced. The phase spectrum of the slowly/rapidly evolving waveform is estimated with the waveform interpolation speech coding model depending on the multiband surd /sonant flag. Then the partial characteristic waveform alignment is used in the speech synthesis section and a 2.4 kbit/s multiband waveform interpolation speech coding algorithm is proposed finally. Compared with the traditional characteristic waveform algorithm, the new algorithm can distinctly improve the speech definition. Compared with the standard 2.4 kbit/s MELP algorithm, the synthetic speech quality is also slightly better.
In the traditional characteristic waveform interpolation speech coding, the high frequency harmonics of the synthetic speech are usually lost, which makes the speech feel noisy, due to the phase information neglect of the characteristic waveform and the whole characteristic waveform alignment. In order to improve the synthetic speech quality, the multiband surd/sonant flag is first introduced. The phase spectrum of the slowly/rapidly evolving waveform is estimated with the waveform interpolation speech coding model depending on the multiband surd /sonant flag. Then the partial characteristic waveform alignment is used in the speech synthesis section and a 2.4 kbit/s multiband waveform interpolation speech coding algorithm is proposed finally. Compared with the traditional characteristic waveform algorithm, the new algorithm can distinctly improve the speech definition. Compared with the standard 2.4 kbit/s MELP algorithm, the synthetic speech quality is also slightly better.
2010, 32(5): 1083-1087.
doi: 10.3724/SP.J.1146.2009.00574
Abstract:
For blind convolutive separation of speech signals, a cost function improved from the least squares fitting function is developed based on second-order statistic, which simplifies the quartic-function with respect to the mixture matrix into three quadratic functions. The iternative approach with three sub-steps is proposed to perform the non-unitary joint block-diagonalization. In each sub-step, a closed solution is derived by minimizing the cost function associated with one parameter-group while fixing the others. Furthermore, the feature of low computational complexity is analytically proven. Compared with ZJBD, the simulations illustrate that the proposed algorithm has the merits of robust initialization selection and better estimate accuracy.
For blind convolutive separation of speech signals, a cost function improved from the least squares fitting function is developed based on second-order statistic, which simplifies the quartic-function with respect to the mixture matrix into three quadratic functions. The iternative approach with three sub-steps is proposed to perform the non-unitary joint block-diagonalization. In each sub-step, a closed solution is derived by minimizing the cost function associated with one parameter-group while fixing the others. Furthermore, the feature of low computational complexity is analytically proven. Compared with ZJBD, the simulations illustrate that the proposed algorithm has the merits of robust initialization selection and better estimate accuracy.
2010, 32(5): 1088-1093.
doi: 10.3724/SP.J.1146.2009.00362
Abstract:
If there are obstacles in the search sea area, the heuristic search algorithm can be applied into the search process of moving underwater targets, to study the heuristic search for moving underwater targets based on Markov process in this paper. The Markov process motion model of underwater targets, the heuristic search model and search probability model of searcher are built. This algorithm continually estimates and updates the moving underwater targets location based on the targets prior location distributed information, to gain accurate targets posterior location distribution information, by using the heuristic function to get the next best search node. The simulation shows that the heuristic search can avoid obstacles effectively, when searching the moving underwater targets. Moreover, it can improve search efficiency. It is useful to study on optimization search for moving underwater targets.
If there are obstacles in the search sea area, the heuristic search algorithm can be applied into the search process of moving underwater targets, to study the heuristic search for moving underwater targets based on Markov process in this paper. The Markov process motion model of underwater targets, the heuristic search model and search probability model of searcher are built. This algorithm continually estimates and updates the moving underwater targets location based on the targets prior location distributed information, to gain accurate targets posterior location distribution information, by using the heuristic function to get the next best search node. The simulation shows that the heuristic search can avoid obstacles effectively, when searching the moving underwater targets. Moreover, it can improve search efficiency. It is useful to study on optimization search for moving underwater targets.
2010, 32(5): 1094-1099.
doi: 10.3724/SP.J.1146.2009.00643
Abstract:
Classical region-based geometric active contours (e.g. C-V model) only take intensity homogeneity as the similarity measure for regions, and can not obtain satisfactory segmentation results of complicated images. Thus, a fast active contour model based on Earth Movers Distance (EMD) is proposed and well adapted to segment images. First, a similarity measure based on EMD is proposed and employed to the segmentation model. Then, a novel regularization and curve evolution method using oversegmentation is enforced to improve the numerical accuracy and evolution efficiency. Experimental results of both synthetic and remote sensing images verify that the algorithm is efficient and accurate.
Classical region-based geometric active contours (e.g. C-V model) only take intensity homogeneity as the similarity measure for regions, and can not obtain satisfactory segmentation results of complicated images. Thus, a fast active contour model based on Earth Movers Distance (EMD) is proposed and well adapted to segment images. First, a similarity measure based on EMD is proposed and employed to the segmentation model. Then, a novel regularization and curve evolution method using oversegmentation is enforced to improve the numerical accuracy and evolution efficiency. Experimental results of both synthetic and remote sensing images verify that the algorithm is efficient and accurate.
2010, 32(5): 1100-1104.
doi: 10.3724/SP.J.1146.2009.00627
Abstract:
Traditional 2D Otsus method supposes that the sum of probabilities of main- diagonal distinct in 2D histogram is approximately one. This papers experiments and theory prove that the sum of probabilities of counter-diagonal distinct in 2D histogram is not always very small and this could not be neglected. Therefore, the assumption mentioned above in 2D Otsus method is inadequately reasonable. In this paper, an improved 2D Otsus segmentation method and recursive algorithm are proposed. By separately calculates probabilities of objects area and background area in 2D histogram main-diagonal distinct, modified method is acquired. Experimental results show that proposed method can obtain better segmentation effect than the traditional 2D Otsus method. The computation complexity of improved 2D Otsus method corresponds to traditional 2D Otsus method.
Traditional 2D Otsus method supposes that the sum of probabilities of main- diagonal distinct in 2D histogram is approximately one. This papers experiments and theory prove that the sum of probabilities of counter-diagonal distinct in 2D histogram is not always very small and this could not be neglected. Therefore, the assumption mentioned above in 2D Otsus method is inadequately reasonable. In this paper, an improved 2D Otsus segmentation method and recursive algorithm are proposed. By separately calculates probabilities of objects area and background area in 2D histogram main-diagonal distinct, modified method is acquired. Experimental results show that proposed method can obtain better segmentation effect than the traditional 2D Otsus method. The computation complexity of improved 2D Otsus method corresponds to traditional 2D Otsus method.
2010, 32(5): 1105-1110.
doi: 10.3724/SP.J.1146.2009.00599
Abstract:
Similarity computation is one of the significant issue in Template-Matching-Method (TMM) based high range resolution automatic target recognition. This paper demonstrates error computation and similarity measurement metrics in similarity computation, and proposes a Relative-Average-Error (RAE) based similarity algorithm employed in TMM. The experiment result shows that, comparing with the widely used Absolute-Mean-Square-Error (AMSE) based TMM, RAE based TMM obtains higher probability of correct classification (Pcc) and more robust recognition ability. Meanwhile, in application, RAE could be an effective solution to take the tradeoff of higher Pcc and less saving memory space for templates.
Similarity computation is one of the significant issue in Template-Matching-Method (TMM) based high range resolution automatic target recognition. This paper demonstrates error computation and similarity measurement metrics in similarity computation, and proposes a Relative-Average-Error (RAE) based similarity algorithm employed in TMM. The experiment result shows that, comparing with the widely used Absolute-Mean-Square-Error (AMSE) based TMM, RAE based TMM obtains higher probability of correct classification (Pcc) and more robust recognition ability. Meanwhile, in application, RAE could be an effective solution to take the tradeoff of higher Pcc and less saving memory space for templates.
2010, 32(5): 1111-1115.
doi: 10.3724/SP.J.1146.2008.01755
Abstract:
For the widely demanding of adaptive multiple moving targets tracking, a type of feature based multi-target tracking algorithm is presented. Background is adaptively modeled by Marr wavelet kernel function and a background subtraction technique based on redundant discrete wavelet transforms is used to detect multiple moving targets. A type of particle filtering combined with SIFT key points is used for tracking, and a queue chain method is used to record data association among different targets, which can improve the detection accuracy and reduce the complexity. Actual tests show that the algorithm can track multi-target with a better performance of real time and mutual occlusion robustness; it can be used in intelligent monitoring with extensive application prospect.
For the widely demanding of adaptive multiple moving targets tracking, a type of feature based multi-target tracking algorithm is presented. Background is adaptively modeled by Marr wavelet kernel function and a background subtraction technique based on redundant discrete wavelet transforms is used to detect multiple moving targets. A type of particle filtering combined with SIFT key points is used for tracking, and a queue chain method is used to record data association among different targets, which can improve the detection accuracy and reduce the complexity. Actual tests show that the algorithm can track multi-target with a better performance of real time and mutual occlusion robustness; it can be used in intelligent monitoring with extensive application prospect.
2010, 32(5): 1116-1120.
doi: 10.3724/SP.J.1146.2009.00298
Abstract:
A new algorithm based on Interacting Multiple Model (IMM) iterated extended Kalman particle filter is proposed for maneuvering target tracking, which uses an improved Particle Filter (PF) in multiple model. First, the Iterated Extended Kalman Filter (IEKF) is modified by providing a new measurement update based on Gauss-Newton iteration, thus the linearity error is reduced. Then the modified IEKF is used to generate the importance density function of PF, which integrates the latest observation into system state transition density. Finally, simulation results are presented to demonstrate the improved performance of the proposed method over Interacting Multiple Model Particle Filter (IMMPF).
A new algorithm based on Interacting Multiple Model (IMM) iterated extended Kalman particle filter is proposed for maneuvering target tracking, which uses an improved Particle Filter (PF) in multiple model. First, the Iterated Extended Kalman Filter (IEKF) is modified by providing a new measurement update based on Gauss-Newton iteration, thus the linearity error is reduced. Then the modified IEKF is used to generate the importance density function of PF, which integrates the latest observation into system state transition density. Finally, simulation results are presented to demonstrate the improved performance of the proposed method over Interacting Multiple Model Particle Filter (IMMPF).
2010, 32(5): 1121-1125.
doi: 10.3724/SP.J.1146.2009.00622
Abstract:
Face recognition is an active research area in the artificial intelligence. A face recognition algorithm using improved Non-negative Matrix Factorization(NMF) with Projected Gradient(PG) for single-trial feature extraction is proposed based on this problem. NMF is a matrix factorization method, which can reflect the inherent partial contact and effectively express single sample information. However, NMF iteration time complexity of the gradient projection optimization method significantly reduces the NMF iteration time complexity of the problem. But the single training sample information has inadequate description of disadvantage, for this disadvantage, before the NMF operator, training sample is filtered by multi-orientation Gabor filters with multi-scale to extract their corresponding local Gabor magnitude map, the PGNMF feature of which were constructed to higher dimensional feature vectors. Experimental results on the ORL face database, YALE face database and FERET face database show that the proposed method is feasible and has higher recognition performance compared with GREY, PCA, ICA, NMF, PGNMF and other algorithms where only one sample image per person is available for training.
Face recognition is an active research area in the artificial intelligence. A face recognition algorithm using improved Non-negative Matrix Factorization(NMF) with Projected Gradient(PG) for single-trial feature extraction is proposed based on this problem. NMF is a matrix factorization method, which can reflect the inherent partial contact and effectively express single sample information. However, NMF iteration time complexity of the gradient projection optimization method significantly reduces the NMF iteration time complexity of the problem. But the single training sample information has inadequate description of disadvantage, for this disadvantage, before the NMF operator, training sample is filtered by multi-orientation Gabor filters with multi-scale to extract their corresponding local Gabor magnitude map, the PGNMF feature of which were constructed to higher dimensional feature vectors. Experimental results on the ORL face database, YALE face database and FERET face database show that the proposed method is feasible and has higher recognition performance compared with GREY, PCA, ICA, NMF, PGNMF and other algorithms where only one sample image per person is available for training.
2010, 32(5): 1126-1130.
doi: 10.3724/SP.J.1146.2009.00549
Abstract:
In this paper a complete on-line handwritten arithmetic expression recognition system is introduced. During the process of stroke grouping and symbol recognition, pre-processing strategy and searching procedure are combined to optimize the confidence function. A novel structural analysis algorithm based on symbol ranking is described, and recognition results are stored in trinary trees as the data structure. Experiments show that relatively neatly written expressions with fairly complex structures can be analysed by this algorithm with a high correction rate.
In this paper a complete on-line handwritten arithmetic expression recognition system is introduced. During the process of stroke grouping and symbol recognition, pre-processing strategy and searching procedure are combined to optimize the confidence function. A novel structural analysis algorithm based on symbol ranking is described, and recognition results are stored in trinary trees as the data structure. Experiments show that relatively neatly written expressions with fairly complex structures can be analysed by this algorithm with a high correction rate.
2010, 32(5): 1131-1135.
doi: 10.3724/SP.J.1146.2009.00748
Abstract:
It is important to identify the source or owner after the vector map is leaked in some key application areas. In this paper proposed scheme, the characteristics of the map owners fingerprint can be extracted and converted into the binary character string as the watermark; and then the watermark will be embedded into the object definition blocks in the attribute file; In any object definition block, there is a character as the ending indicator followed the definition information of the object, the data after the ending character can not be processed or displayed by any tools. Experiments show that the algorithm has better robustness; The watermark can be embedded or extracted nondestructively to the vector map, the value of NC coefficient is more than 97%, as the factor of fingerprint matching algorithm, the accuracy of authentication is more than 80%.
It is important to identify the source or owner after the vector map is leaked in some key application areas. In this paper proposed scheme, the characteristics of the map owners fingerprint can be extracted and converted into the binary character string as the watermark; and then the watermark will be embedded into the object definition blocks in the attribute file; In any object definition block, there is a character as the ending indicator followed the definition information of the object, the data after the ending character can not be processed or displayed by any tools. Experiments show that the algorithm has better robustness; The watermark can be embedded or extracted nondestructively to the vector map, the value of NC coefficient is more than 97%, as the factor of fingerprint matching algorithm, the accuracy of authentication is more than 80%.
2010, 32(5): 1136-1140.
doi: 10.3724/SP.J.1146.2009.00530
Abstract:
Secret message can be embedded into letters in tags of a webpage in ways that are imperceptible to human eye viewed with a browser. These messages, however, alter the inherent characteristic of the offset of a tag. This paper presents a new higher-order statistical steganalytic algorithm for detection of secret messages embedded in a webpage. The offset is used to build the higher-order statistical models to detect whether secret messages hide in tags. 30 homepages are randomly downloaded from different websites to test, and the results show the reliability and accuracy of statistical characteristics. The probability of missing secret messages decrease as the secret message increase, and it is zero, as 50% letters of tags are used to carry secret message.
Secret message can be embedded into letters in tags of a webpage in ways that are imperceptible to human eye viewed with a browser. These messages, however, alter the inherent characteristic of the offset of a tag. This paper presents a new higher-order statistical steganalytic algorithm for detection of secret messages embedded in a webpage. The offset is used to build the higher-order statistical models to detect whether secret messages hide in tags. 30 homepages are randomly downloaded from different websites to test, and the results show the reliability and accuracy of statistical characteristics. The probability of missing secret messages decrease as the secret message increase, and it is zero, as 50% letters of tags are used to carry secret message.
2010, 32(5): 1141-1145.
doi: 10.3724/SP.J.1146.2009.00641
Abstract:
It is a valuable task of mining hot topics on Internet utilizing users feedback under Web 2.0 environment. Motivated by heat diffusion phenomena, the information activities on the Internet are treated as heat flow, and heat diffusion model is used to simulate these acitivities. Based on heat diffusion model, a novel topic detection and ranking algorithm is proposed. The experiment results demonstrate that the algorithm works well under Web 2.0 environment.
It is a valuable task of mining hot topics on Internet utilizing users feedback under Web 2.0 environment. Motivated by heat diffusion phenomena, the information activities on the Internet are treated as heat flow, and heat diffusion model is used to simulate these acitivities. Based on heat diffusion model, a novel topic detection and ranking algorithm is proposed. The experiment results demonstrate that the algorithm works well under Web 2.0 environment.
2010, 32(5): 1146-1151.
doi: 10.3724/SP.J.1146.2009.00523
Abstract:
In non-cooperative communication scenarios, coherent demodulation of Continuous Phase Modulation (CPM) signals requires precise estimation of CPM modulation parameters. A new data-aided modulation index and shaping pulse joint estimation algorithm is proposed for CPM signals. The differential phase between symbols is used to eliminate signal phase memories, and Least-Square method is used for parameters estimations. Theoretical analysis and numerical simulations show that it is unbiased. The estimating error is converged to zero at 1/N where N is the length of observation data. The algorithm is applicable to both full and partial response CPM modulations and presents good anti-noise performance.
In non-cooperative communication scenarios, coherent demodulation of Continuous Phase Modulation (CPM) signals requires precise estimation of CPM modulation parameters. A new data-aided modulation index and shaping pulse joint estimation algorithm is proposed for CPM signals. The differential phase between symbols is used to eliminate signal phase memories, and Least-Square method is used for parameters estimations. Theoretical analysis and numerical simulations show that it is unbiased. The estimating error is converged to zero at 1/N where N is the length of observation data. The algorithm is applicable to both full and partial response CPM modulations and presents good anti-noise performance.
2010, 32(5): 1152-1156.
doi: 10.3724/SP.J.1146.2009.00838
Abstract:
Most of the proposed constructions for Low-Density Parity-Check (LDPC) codes with girth at least eight are focused on (semi-) stochastic methods with the aid of computer search. Motivated by the resulting parity-check matrices obtained from the Greedy construction idea, a deterministic method is presented to construct a novel family of girth-8 (3,L)-regular Quasi-Cyclic (QC-) LDPC codes. The parity-check matrix of the new code consists of 3L PP cyclic permutation matrices and the girth of its Tanner graph is eight for arbitrary integer P3L2/4.
Most of the proposed constructions for Low-Density Parity-Check (LDPC) codes with girth at least eight are focused on (semi-) stochastic methods with the aid of computer search. Motivated by the resulting parity-check matrices obtained from the Greedy construction idea, a deterministic method is presented to construct a novel family of girth-8 (3,L)-regular Quasi-Cyclic (QC-) LDPC codes. The parity-check matrix of the new code consists of 3L PP cyclic permutation matrices and the girth of its Tanner graph is eight for arbitrary integer P3L2/4.
2010, 32(5): 1157-1162.
doi: 10.3724/SP.J.1146.2009.00586
Abstract:
For Space-Time Block Code (STBC) systems with unknown co-channel interference in multi-cell environment, focusing on the problem that the Capon-style receivers will result the self-interference in the STBC codeword from the finite samples, a robust linear receiver is proposed. Based on the orthogonal structure of the STBC, the self-interference in the STBC codeword is first canceled with the orthogonal projection. Then Capon receiver is used to suppress the unknown interference. Simulation results are presented to demonstrate the significant gain obtained by the proposed linear receivers over the conventional Capon receiver.
For Space-Time Block Code (STBC) systems with unknown co-channel interference in multi-cell environment, focusing on the problem that the Capon-style receivers will result the self-interference in the STBC codeword from the finite samples, a robust linear receiver is proposed. Based on the orthogonal structure of the STBC, the self-interference in the STBC codeword is first canceled with the orthogonal projection. Then Capon receiver is used to suppress the unknown interference. Simulation results are presented to demonstrate the significant gain obtained by the proposed linear receivers over the conventional Capon receiver.
2010, 32(5): 1163-1167.
doi: 10.3724/SP.J.1146.2009.00614
Abstract:
For Multiple-Input Multiple-Output with Orthogonal Frequency Division Multiplexing (MIMO-OFDM) systems, Maximum A posteriori Probability (MAP) channel estimation algorithm uses expectation maximum (EM) algorithm to reduce complexity, but it will lead to error floor phenomenon. The data transmission efficiency of systems is subject to the number of transmit antennas. According to these problems, in this paper, an effective MAP algorithm is proposed and its performance is analyzed. The proposed algorithm lowers the estimated error depending on the spatial independence of MIMO channels in angle domain on the basis of making use of EM algorithm to decrease the complexity. To improve the data transmission efficiency and the performance of estimation, joint estimation is carried out over multiple OFDM symbols. Simulation results indicate the proposed algorithm obtains better estimated performance and higher data transmission efficiency.
For Multiple-Input Multiple-Output with Orthogonal Frequency Division Multiplexing (MIMO-OFDM) systems, Maximum A posteriori Probability (MAP) channel estimation algorithm uses expectation maximum (EM) algorithm to reduce complexity, but it will lead to error floor phenomenon. The data transmission efficiency of systems is subject to the number of transmit antennas. According to these problems, in this paper, an effective MAP algorithm is proposed and its performance is analyzed. The proposed algorithm lowers the estimated error depending on the spatial independence of MIMO channels in angle domain on the basis of making use of EM algorithm to decrease the complexity. To improve the data transmission efficiency and the performance of estimation, joint estimation is carried out over multiple OFDM symbols. Simulation results indicate the proposed algorithm obtains better estimated performance and higher data transmission efficiency.
A Minimum Jamming Precoding Algorithm and Its Performance Analysis for the Downlink Multiuser System
2010, 32(5): 1168-1172.
doi: 10.3724/SP.J.1146.2009.00623
Abstract:
In this paper, a linearly precoding algorithm is presented based on the minimum jamming criterion for the downlink multi-element multi-user system. And by using nonlinear programming method, the closed-form expression of precoding weights is also derived, therefore reducing the complexity of the system significantly. Furthermore, due to its interference-limited property, the exact Probability Density Function (PDF) of the Signal-to-Interference Ratio (SIR) is also derived via the closed-form expression obtained, thus facilitating the analysis and application of the algorithm.
In this paper, a linearly precoding algorithm is presented based on the minimum jamming criterion for the downlink multi-element multi-user system. And by using nonlinear programming method, the closed-form expression of precoding weights is also derived, therefore reducing the complexity of the system significantly. Furthermore, due to its interference-limited property, the exact Probability Density Function (PDF) of the Signal-to-Interference Ratio (SIR) is also derived via the closed-form expression obtained, thus facilitating the analysis and application of the algorithm.
2010, 32(5): 1173-1177.
doi: 10.3724/SP.J.1146.2009.00783
Abstract:
The small spacing between the double circular polarized antennas installed in the small mobile terminal can lead to the strong spatial correlation, which will directly diminish the channel capacity. From the theoretical point, this paper studies the relationship between the spatial correlation and the parameters which include the reconfigurable antenna parameters (such as the polarization rotating direction, the polarization axial ratio) and environment parameters (such as the channel cross polarization discrimination, the arrival angle). This paper also studies how to decrease the spatial correlation by adjusting the above parameters. Based on the analysis and simulation results, the design method of adaptive reconfiguration antenna is presented. The design method proposed in the paper can be used as reference for the design of reconfigurable polarization antenna in the small MIMO mobile terminal.
The small spacing between the double circular polarized antennas installed in the small mobile terminal can lead to the strong spatial correlation, which will directly diminish the channel capacity. From the theoretical point, this paper studies the relationship between the spatial correlation and the parameters which include the reconfigurable antenna parameters (such as the polarization rotating direction, the polarization axial ratio) and environment parameters (such as the channel cross polarization discrimination, the arrival angle). This paper also studies how to decrease the spatial correlation by adjusting the above parameters. Based on the analysis and simulation results, the design method of adaptive reconfiguration antenna is presented. The design method proposed in the paper can be used as reference for the design of reconfigurable polarization antenna in the small MIMO mobile terminal.
2010, 32(5): 1178-1182.
doi: 10.3724/SP.J.1146.2009.00587
Abstract:
Intrinsic Time-scale Decomposition(ITD) algorithm is a kind of local wave analysis algorithm. In this paper, Direct Sequence Spread Spectrum (DSSS) signal is decomposed by ITD and a fast algorithm for detecting the chip rate and carrier frequency of direct spread spectrum signal is proposed according to coarse search and elaborate search in frequency domain separately. This method is based on the parameter of instant amplitude,where the signal is firstly low-pass filtered by a cut-off frequency and the noise is then eliminated by an incidental frequency. High time-frequency resolution and fast calculation speed of ITD algorithm are used in this algorithm. Simulation results show that this algorithm can detect chip rate and carrier frequency effectively when SNR is of -15dB.
Intrinsic Time-scale Decomposition(ITD) algorithm is a kind of local wave analysis algorithm. In this paper, Direct Sequence Spread Spectrum (DSSS) signal is decomposed by ITD and a fast algorithm for detecting the chip rate and carrier frequency of direct spread spectrum signal is proposed according to coarse search and elaborate search in frequency domain separately. This method is based on the parameter of instant amplitude,where the signal is firstly low-pass filtered by a cut-off frequency and the noise is then eliminated by an incidental frequency. High time-frequency resolution and fast calculation speed of ITD algorithm are used in this algorithm. Simulation results show that this algorithm can detect chip rate and carrier frequency effectively when SNR is of -15dB.
2010, 32(5): 1183-1187.
doi: 10.3724/SP.J.1146.2009.00556
Abstract:
A virtual MIMO transmission scheme based on Discrete Fourier Transform Spread Orthogonal Frequency Division Multiplexing (DFT-S-OFDM) with Block-Space-Time Block Coding (B-STBC)is proposed and a lower complexity detection algorithm for the virtual MIMO system is also presented in this paper. Simulation results show that the proposed scheme can keep the same Peak to Average Power Ratio (PAPR) characteristic as the original DFT-S-OFDM system and reduce the detection threshold of virtual MIMO system, which can improve the transmission reliability in the uplink virtual MIMO system.
A virtual MIMO transmission scheme based on Discrete Fourier Transform Spread Orthogonal Frequency Division Multiplexing (DFT-S-OFDM) with Block-Space-Time Block Coding (B-STBC)is proposed and a lower complexity detection algorithm for the virtual MIMO system is also presented in this paper. Simulation results show that the proposed scheme can keep the same Peak to Average Power Ratio (PAPR) characteristic as the original DFT-S-OFDM system and reduce the detection threshold of virtual MIMO system, which can improve the transmission reliability in the uplink virtual MIMO system.
2010, 32(5): 1188-1192.
doi: 10.3724/SP.J.1146.2009.00463
Abstract:
This paper proposes a novel transform coding method based on the Karhunen-Loeve (KL) transform in multiple linear spaces. This method uses multiple linear subspaces to approximate image signals, and use a KL transform for each linear subspace, so that the transform coefficients in each subspace is close to Gaussian, which effectively de-correlates the coefficients. An SNR-scalable video coding scheme is developed by combining the method with wavelet decomposition. Experiments prove that this scheme accomplishes better coding efficiency than the schemes employing DCT, DWT, and Directional DCT methods.
This paper proposes a novel transform coding method based on the Karhunen-Loeve (KL) transform in multiple linear spaces. This method uses multiple linear subspaces to approximate image signals, and use a KL transform for each linear subspace, so that the transform coefficients in each subspace is close to Gaussian, which effectively de-correlates the coefficients. An SNR-scalable video coding scheme is developed by combining the method with wavelet decomposition. Experiments prove that this scheme accomplishes better coding efficiency than the schemes employing DCT, DWT, and Directional DCT methods.
2010, 32(5): 1193-1199.
doi: 10.3724/SP.J.1146.2009.00775
Abstract:
Due to the deficiencies of current schemes in the aspecct of energy consumption efficiently, this paper proposes a novel Multi-Radio Channel Searching Scheme (MRCSS) based on Partially Observable Markov Decision Process (POMDP) in rayleigh fading environment. This scheme can obtain the optimum energy-efficient channel through analysis and derivation of channel state and thereby conduct unliecense users channel selection. The simulation results show the proposed scheme can reduce the searching time and save the energy more efficiently than the classical methods, so as to improve the spectrum sensing performance.
Due to the deficiencies of current schemes in the aspecct of energy consumption efficiently, this paper proposes a novel Multi-Radio Channel Searching Scheme (MRCSS) based on Partially Observable Markov Decision Process (POMDP) in rayleigh fading environment. This scheme can obtain the optimum energy-efficient channel through analysis and derivation of channel state and thereby conduct unliecense users channel selection. The simulation results show the proposed scheme can reduce the searching time and save the energy more efficiently than the classical methods, so as to improve the spectrum sensing performance.
2010, 32(5): 1200-1204.
doi: 10.3724/SP.J.1146.2009.00744
Abstract:
Cooperative communication enables nodes to share their antennas to form a virtual multiple-input multiple-output system and achieve spatial diversity gain. Exploiting the diversity gain of V-MIMO channels, a multi-packet reception method is proposed to employ cooperative strategies in up links of wireless mesh networks. The method can support random amount of collided mesh clients by selecting cooperative nodes to form V-MIMO channels. Mesh routers with multiple antennas adopt the successive interference cancellation algorithm to receive separate the collided packets quickly. The theoretical analysis and simulation results show that the proposed method is capable of fully utilizing the spatial resources and outperforms the conventional NDMA and TDMA in terms of system throughput and transmission delay.
Cooperative communication enables nodes to share their antennas to form a virtual multiple-input multiple-output system and achieve spatial diversity gain. Exploiting the diversity gain of V-MIMO channels, a multi-packet reception method is proposed to employ cooperative strategies in up links of wireless mesh networks. The method can support random amount of collided mesh clients by selecting cooperative nodes to form V-MIMO channels. Mesh routers with multiple antennas adopt the successive interference cancellation algorithm to receive separate the collided packets quickly. The theoretical analysis and simulation results show that the proposed method is capable of fully utilizing the spatial resources and outperforms the conventional NDMA and TDMA in terms of system throughput and transmission delay.
2010, 32(5): 1205-1211.
doi: 10.3724/SP.J.1146.2009.00201
Abstract:
The efficiency of energy consumption in wireless sensor networks is closely related to the traffic of each node. In this paper, the issue of load balancing in wireless sensor networks is researched on the microscopic scale. Based on an equiprobable routing model, the relationship between topology structure and data traffic distribution is established. Based on the analytical results, the definition and the distributed algorithm of load density of any node in many-to-one sensor networks are proposed. Based on the first order radio model, the relationship between load density and network lifetime is analyzed. In spite of the fact that complete load balancing in wireless sensor network is unreachable, suboptimal load balancing is possible if the topology structure is well-designed. Simulation results show that the load distribution of wireless sensor networks can be described accurately by load density, and load balancing of most nodes is possible if topology structure meets suboptimal conditions.
The efficiency of energy consumption in wireless sensor networks is closely related to the traffic of each node. In this paper, the issue of load balancing in wireless sensor networks is researched on the microscopic scale. Based on an equiprobable routing model, the relationship between topology structure and data traffic distribution is established. Based on the analytical results, the definition and the distributed algorithm of load density of any node in many-to-one sensor networks are proposed. Based on the first order radio model, the relationship between load density and network lifetime is analyzed. In spite of the fact that complete load balancing in wireless sensor network is unreachable, suboptimal load balancing is possible if the topology structure is well-designed. Simulation results show that the load distribution of wireless sensor networks can be described accurately by load density, and load balancing of most nodes is possible if topology structure meets suboptimal conditions.
2010, 32(5): 1212-1218.
doi: 10.3724/SP.J.1146.2009.00295
Abstract:
For that the communications by unreliable links wastes too much energy in wireless sensor networks, based on link quality, routing cost functions are constructed and proper selective probability is allocated to paths. Then the algorithm CBLQ (Cost Balanced routing algorithm based on Link Quality) is designed by adopting the minimum-hop-count forwarding strategy. To further reduce and balance the routing costs, the alternate forwarding condition of brother nodes is analyzed and the new routing algorithm, CBLQ-E, is extended to form. The results of simulating experiments prove that both algorithms increase the ratio of energy utilization effectively, as well as decrease the data transmission delay of networks.
For that the communications by unreliable links wastes too much energy in wireless sensor networks, based on link quality, routing cost functions are constructed and proper selective probability is allocated to paths. Then the algorithm CBLQ (Cost Balanced routing algorithm based on Link Quality) is designed by adopting the minimum-hop-count forwarding strategy. To further reduce and balance the routing costs, the alternate forwarding condition of brother nodes is analyzed and the new routing algorithm, CBLQ-E, is extended to form. The results of simulating experiments prove that both algorithms increase the ratio of energy utilization effectively, as well as decrease the data transmission delay of networks.
2010, 32(5): 1219-1223.
doi: 10.3724/SP.J.1146.2009.00754
Abstract:
The proposed proxy re-signature schemes had no message blinded properties as yet, so there is a problem that they can not protect the original signers privacy. The general proxy re-signature scheme is extended using the idea of blind signature. The definition of blind proxy re-signature scheme been given firstly; then according to the definition, the blind proxy re-signatures scheme Based on Standard Model is proposed using of bilinear mapping technology and standard model framework proposed by Waters, finally, the security of the scheme is proven. The transparent signature conversion from original signer to proxy re-signer is achieved in this paper which can protect the original signers privacy, at the same time, the scheme is practicality as it is designed using the framework of standard model.
The proposed proxy re-signature schemes had no message blinded properties as yet, so there is a problem that they can not protect the original signers privacy. The general proxy re-signature scheme is extended using the idea of blind signature. The definition of blind proxy re-signature scheme been given firstly; then according to the definition, the blind proxy re-signatures scheme Based on Standard Model is proposed using of bilinear mapping technology and standard model framework proposed by Waters, finally, the security of the scheme is proven. The transparent signature conversion from original signer to proxy re-signer is achieved in this paper which can protect the original signers privacy, at the same time, the scheme is practicality as it is designed using the framework of standard model.
2010, 32(5): 1224-1228.
doi: 10.3724/SP.J.1146.2009.00739
Abstract:
MTL (Multiconductor Transmission Line) inductance matrix is commonly computed through complex and tiresome indirect transform method, so a novel method to directly compute the MTL inductance matrix is proposed. Double filaments circuit is utilized to construct a matrix model, and then inductance matrix is directly solved from the model by the novel method, so the complexity of the analysis is greatly reduced. Additionally, matrix transformation is adopted; therefore, currents distribution can also be supplied when inductance matrix calculation is finished, which is impossible for classical indirect method. The novel method imports nonmagnetic quasi-static hypothesis and filaments division, then it can be applied to the transmission line structure with arbitrary shaped cross-section, and arbitrary separate distance in wide frequency band. Simulations show that this method has a good precision in currents distribution and inductance matrix calculation.
MTL (Multiconductor Transmission Line) inductance matrix is commonly computed through complex and tiresome indirect transform method, so a novel method to directly compute the MTL inductance matrix is proposed. Double filaments circuit is utilized to construct a matrix model, and then inductance matrix is directly solved from the model by the novel method, so the complexity of the analysis is greatly reduced. Additionally, matrix transformation is adopted; therefore, currents distribution can also be supplied when inductance matrix calculation is finished, which is impossible for classical indirect method. The novel method imports nonmagnetic quasi-static hypothesis and filaments division, then it can be applied to the transmission line structure with arbitrary shaped cross-section, and arbitrary separate distance in wide frequency band. Simulations show that this method has a good precision in currents distribution and inductance matrix calculation.
2010, 32(5): 1229-1233.
doi: 10.3724/SP.J.1146.2009.00631
Abstract:
A innovative digital signal processing technique for fast measurements of ElectroMagnetic Interference(EMI) from devices under test in the time-domain at open area test sites is presented. Suppressing ambient noise by using the wavelet analysis, the EMI measurements of an Equipment Under Test(EUT) can be performed at a test site polluted with electromagnetic ambient noise. Frequency-domain threshold funtion filtering using the wavelet analysis method is applied. The coherence of ambient noise is computed by using the wavelet analysis in time domain and frequency domain. With the characteristic of ambient noise, it can be suppressed by the threshold function in the wavelet domain, and the pure electromagnetic radiation can be obtained readily. The EMI measurement can be performed without the anechoic chamber. It is seen that it is convenient to perform the EMI measurements while the total cost can be reduced greatly. Measurement results show that the successful cancelation of ambient noise can be obtained in the frequency range of 30 Hz to 1000 MHz.
A innovative digital signal processing technique for fast measurements of ElectroMagnetic Interference(EMI) from devices under test in the time-domain at open area test sites is presented. Suppressing ambient noise by using the wavelet analysis, the EMI measurements of an Equipment Under Test(EUT) can be performed at a test site polluted with electromagnetic ambient noise. Frequency-domain threshold funtion filtering using the wavelet analysis method is applied. The coherence of ambient noise is computed by using the wavelet analysis in time domain and frequency domain. With the characteristic of ambient noise, it can be suppressed by the threshold function in the wavelet domain, and the pure electromagnetic radiation can be obtained readily. The EMI measurement can be performed without the anechoic chamber. It is seen that it is convenient to perform the EMI measurements while the total cost can be reduced greatly. Measurement results show that the successful cancelation of ambient noise can be obtained in the frequency range of 30 Hz to 1000 MHz.
2010, 32(5): 1234-1238.
doi: 10.3724/SP.J.1146.2009.01223
Abstract:
Compressive sensing technique has been shown to be able to reduce effectively the data rate of high- resolution radar imaging system. A phase-preserving range compression algorithm based on Compressive Sensing (CS) technique is proposed, after analyzing the sparse characteristics of complex base-band echo signal from a target using chirp signal as transmitted signal. Based on this range compression algorithm, a 2D imaging scheme is established, i.e. performing range compression by using CS technique and performing azimuth compression by using traditional technique. The effectiveness of the 2D imaging scheme is tested through processing both simulated data and real radar data.
Compressive sensing technique has been shown to be able to reduce effectively the data rate of high- resolution radar imaging system. A phase-preserving range compression algorithm based on Compressive Sensing (CS) technique is proposed, after analyzing the sparse characteristics of complex base-band echo signal from a target using chirp signal as transmitted signal. Based on this range compression algorithm, a 2D imaging scheme is established, i.e. performing range compression by using CS technique and performing azimuth compression by using traditional technique. The effectiveness of the 2D imaging scheme is tested through processing both simulated data and real radar data.
2010, 32(5): 1239-1243.
doi: 10.3724/SP.J.1146.2009.00565
Abstract:
The sub-aperture algorithm has great important role in SAR real-time processing. The algorithm combined the sub-aperture and Nonlinear Chirp Scaling (NCS) algorithm has good performance in real-time processing of UWB SAR. However the ghost or false targets exist in the algorithm based on sub-aperture. In this paper, the reason for ghost or false target existing is theoretically analyzed, furthermore, the method of eliminating the ghost is proposed using zero-padding with sub-aperture data. Finally, the validity is verified with simulation and real UWB SAR data.
The sub-aperture algorithm has great important role in SAR real-time processing. The algorithm combined the sub-aperture and Nonlinear Chirp Scaling (NCS) algorithm has good performance in real-time processing of UWB SAR. However the ghost or false targets exist in the algorithm based on sub-aperture. In this paper, the reason for ghost or false target existing is theoretically analyzed, furthermore, the method of eliminating the ghost is proposed using zero-padding with sub-aperture data. Finally, the validity is verified with simulation and real UWB SAR data.
2010, 32(5): 1244-1247.
doi: 10.3724/SP.J.1146.2009.00640
Abstract:
Interferometric parameters calibration of InSAR system can acquire both interferometric phase bias and accurate baseline parameters. Then planimetric and vertical accuracies can be improved much. An approach of interferometric calibration over very large areas is presented in this paper. It acquired corresponding points using feature points extraction and optimal matching, then weighted bundle adjustment is induced to realize calibration of each sub-data simultaneously. The calibration parameters are analyzed. Results using airborne data show robust of the algorithm with respect to different terrain type.
Interferometric parameters calibration of InSAR system can acquire both interferometric phase bias and accurate baseline parameters. Then planimetric and vertical accuracies can be improved much. An approach of interferometric calibration over very large areas is presented in this paper. It acquired corresponding points using feature points extraction and optimal matching, then weighted bundle adjustment is induced to realize calibration of each sub-data simultaneously. The calibration parameters are analyzed. Results using airborne data show robust of the algorithm with respect to different terrain type.
2010, 32(5): 1248-1252.
doi: 10.3724/SP.J.1146.2009.00544
Abstract:
In this paper a new kind of waveform is proposed to improve covertness of active sonar. A specified code is used to modulate the chaotic signal as waveform of active sonar. The issues of the distortions introduced by filtered by physical underwater channel and additive noise are discussed .Echoes are demodulated by adaptive filter .Recovered signals are identified by stored specified codes. The Lyapunov adaptive filter improved in phase space is applied to be demodulator. The obtained results indicate that proposed demodulator quickly and accurately recover original chaotic signal distorted by additive noise and filter. It is found that chaotic waveforms are better covertness than the classical waveforms of active sonar.
In this paper a new kind of waveform is proposed to improve covertness of active sonar. A specified code is used to modulate the chaotic signal as waveform of active sonar. The issues of the distortions introduced by filtered by physical underwater channel and additive noise are discussed .Echoes are demodulated by adaptive filter .Recovered signals are identified by stored specified codes. The Lyapunov adaptive filter improved in phase space is applied to be demodulator. The obtained results indicate that proposed demodulator quickly and accurately recover original chaotic signal distorted by additive noise and filter. It is found that chaotic waveforms are better covertness than the classical waveforms of active sonar.
2010, 32(5): 1253-1256.
doi: 10.3724/SP.J.1146.2009.00460
Abstract:
A novel method of multi-sources distinguishing is proposed single acoustic vector sensor by detecting the line spectra. The distinguishing capability is found to be only relative to the output SNR of the line spectra. with the direction finding ability of acoustic vector sensor with in low frequency band, the association between the line spectra and the corresponding sources is established by using direction vicinity principle, the multi-sources distinguishing and auto-tracking is then realized with filtering. Simulations with sea trial data show that adjacent 3 sources with in 10 can be distinguished with single acoustic vector sensor. This method is simple and effective, thus has the prospect of engineering applications.
A novel method of multi-sources distinguishing is proposed single acoustic vector sensor by detecting the line spectra. The distinguishing capability is found to be only relative to the output SNR of the line spectra. with the direction finding ability of acoustic vector sensor with in low frequency band, the association between the line spectra and the corresponding sources is established by using direction vicinity principle, the multi-sources distinguishing and auto-tracking is then realized with filtering. Simulations with sea trial data show that adjacent 3 sources with in 10 can be distinguished with single acoustic vector sensor. This method is simple and effective, thus has the prospect of engineering applications.
2010, 32(5): 1257-1261.
doi: 10.3724/SP.J.1146.2009.00726
Abstract:
The average Hamming correlation is an important performance indicator of the frequency hopping sequences. Based on power residue module p, a class of frequency-hopping sequence family with length of sequences being p2 and family size (p-1)2 is constructed in this paper. It is shown that the average Hamming autocorrelation of the new frequency hopping sequence family is 0, and the average Hamming crosscorrelation is 1. The family is optimal with respect to the average Hamming correlation bound.
The average Hamming correlation is an important performance indicator of the frequency hopping sequences. Based on power residue module p, a class of frequency-hopping sequence family with length of sequences being p2 and family size (p-1)2 is constructed in this paper. It is shown that the average Hamming autocorrelation of the new frequency hopping sequence family is 0, and the average Hamming crosscorrelation is 1. The family is optimal with respect to the average Hamming correlation bound.
2010, 32(5): 1262-1266.
doi: 10.3724/SP.J.1146.2009.00647
Abstract:
In the 802.16e system, when multiple Mobile Stations (MS) with real-time services enter sleep mode, the efficiency of power saving is impacted by the competition of the data scheduling in the overlapping listening windows. This paper analyzes the impaction and proposes an enhanced sleep mode management scheme. The proposed scheme tries to make the listening windows of different MS lie dispersedly and balance the system load among OFDM frames by adjusting the start time of sleep mode of the new connection. The simulation result shows the proposed scheme not only improves the energy efficiency but also decreases the signaling overhead as well as improves the QoS of the network.
In the 802.16e system, when multiple Mobile Stations (MS) with real-time services enter sleep mode, the efficiency of power saving is impacted by the competition of the data scheduling in the overlapping listening windows. This paper analyzes the impaction and proposes an enhanced sleep mode management scheme. The proposed scheme tries to make the listening windows of different MS lie dispersedly and balance the system load among OFDM frames by adjusting the start time of sleep mode of the new connection. The simulation result shows the proposed scheme not only improves the energy efficiency but also decreases the signaling overhead as well as improves the QoS of the network.
2010, 32(5): 1267-1270.
doi: 10.3724/SP.J.1146.2009.00568
Abstract:
The influence of Electromagnetic Band-Gap (EBG) structure on signal transmission is investigated in this paper. The signal integrity in time domain and frequency domain is analyzed by HFSS and S parameter model. The simulation results show that the signal reflection and filter are induced by periodic structure, signal quality is affected seriously .The smaller period is, the worse signal quality is. Finally, the design rules for high-speed circuit in EBG structure are derived from the results obtained.
The influence of Electromagnetic Band-Gap (EBG) structure on signal transmission is investigated in this paper. The signal integrity in time domain and frequency domain is analyzed by HFSS and S parameter model. The simulation results show that the signal reflection and filter are induced by periodic structure, signal quality is affected seriously .The smaller period is, the worse signal quality is. Finally, the design rules for high-speed circuit in EBG structure are derived from the results obtained.
