Email alert
2012 Vol. 34, No. 6
Display Method:
2012, 34(6): 1271-1276.
doi: 10.3724/SP.J.1146.2011.01108
Abstract:
To solve the issue of high false-alarm rate and the target missing effect in frequency ambiguous areas during the detection process, a CFAR localization algorithm based on the high dimensional frequency fitting technology is proposed in this paper. For this algorithm, the CFAR technology is first applied to real-time adjust the system noise threshold and suppress the rising of the fake targets. Then it takes advantage of the two-dimensional fitting process to realize the preliminary recognition and estimation of the components. Finally, based on the practical demands of the detection, the adaptive frequency fitting algorithm is used to modify the coefficients of the identified components and locate the targets accurately. Experimental results prove that the proposed new algorithm can effectively eliminate the false targets, improve the system adaptability to different detection environments and precisely identify multiple targets even in frequency ambiguous area.
To solve the issue of high false-alarm rate and the target missing effect in frequency ambiguous areas during the detection process, a CFAR localization algorithm based on the high dimensional frequency fitting technology is proposed in this paper. For this algorithm, the CFAR technology is first applied to real-time adjust the system noise threshold and suppress the rising of the fake targets. Then it takes advantage of the two-dimensional fitting process to realize the preliminary recognition and estimation of the components. Finally, based on the practical demands of the detection, the adaptive frequency fitting algorithm is used to modify the coefficients of the identified components and locate the targets accurately. Experimental results prove that the proposed new algorithm can effectively eliminate the false targets, improve the system adaptability to different detection environments and precisely identify multiple targets even in frequency ambiguous area.
2012, 34(6): 1277-1283.
doi: 10.3724/SP.J.1146.2011.00935
Abstract:
Through-wall imaging techniques with Ultra-WideBand (UWB) radar are promising candidates for non-destructive detection, especially in disaster areas like calamity rescue scenarios, counter-terrorism and so on. These applications always require high-resolution target imaging to identify the object shape. Based on this requirement, a complex target boundary imaging algorithm for UWB radar is proposed to realize the shape estimation and identification of the target behind walls. As the application is a near field problem, the algorithm offsets the influence from walls using the incident angle of the radar echo calculated by the time-delay difference curve, and estimates the target boundary effectively. Simulation and experimental results of targets behind walls with different shapes show that the proposed algorithm can remove the influence from walls, estimate the target boundary effectively and realize the basic target identification.
Through-wall imaging techniques with Ultra-WideBand (UWB) radar are promising candidates for non-destructive detection, especially in disaster areas like calamity rescue scenarios, counter-terrorism and so on. These applications always require high-resolution target imaging to identify the object shape. Based on this requirement, a complex target boundary imaging algorithm for UWB radar is proposed to realize the shape estimation and identification of the target behind walls. As the application is a near field problem, the algorithm offsets the influence from walls using the incident angle of the radar echo calculated by the time-delay difference curve, and estimates the target boundary effectively. Simulation and experimental results of targets behind walls with different shapes show that the proposed algorithm can remove the influence from walls, estimate the target boundary effectively and realize the basic target identification.
2012, 34(6): 1284-1290.
doi: 10.3724/SP.J.1146.2011.00726
Abstract:
Inverse Synthetic Aperture LADAR (ISAL) can realize high resolution imaging for targets. However, if there are rotating parts in the targets, the echo from rotating parts would influence the imaging for targets. A micro-Doppler feature extraction and imaging method for the targets with rotating parts in ISAL is proposed. The echo is compensated by matched reference signal method firstly. Then, the micro-Doppler curves are separated from frequency-slow time spectrogram. The parameters of micro-Doppler curves are extracted fast based on the periodicity of the curves. Finally, the echo of rotating parts is separate from the echo of target body, and the 2-D imaging for target is realized. Simulation shows that the proposed method can not only filter out the pollution caused by rotating parts in ISAL imaging, but also provide new information for the target identification.
Inverse Synthetic Aperture LADAR (ISAL) can realize high resolution imaging for targets. However, if there are rotating parts in the targets, the echo from rotating parts would influence the imaging for targets. A micro-Doppler feature extraction and imaging method for the targets with rotating parts in ISAL is proposed. The echo is compensated by matched reference signal method firstly. Then, the micro-Doppler curves are separated from frequency-slow time spectrogram. The parameters of micro-Doppler curves are extracted fast based on the periodicity of the curves. Finally, the echo of rotating parts is separate from the echo of target body, and the 2-D imaging for target is realized. Simulation shows that the proposed method can not only filter out the pollution caused by rotating parts in ISAL imaging, but also provide new information for the target identification.
2012, 34(6): 1291-1296.
doi: 10.3724/SP.J.1146.2011.00666
Abstract:
One of the main objectives of High Frequency Surface Wave Radar (HFSWR) is to adapt the spectrum of transmit waveform to avoid narrowband interferences of wireless communication systems. Sparse frequency waveform that contains several individual clear bands and stop bands is a desirable solution to this problem. Besides spectral requirements, good correlation properties of transmit waveform are also desired in specific applications, such as the range side lobe. Moreover, practical hardware constraints require real-time algorithm which has a fast convergence rate. In this paper, a new penalty function is proposed based on both requirements for the Power Spectrum Density (PSD) and the Integrated Sidelobe Level (ISL). Then, the phase-only conjugate gradient method with very fast convergence rate is proposed to solve this problem. The proposed approach is efficient in computation and flexible in designing waveform. Optimized waveform can effectively suppress narrowband interference and have low ISL. The validity of the proposed method is illustrated by numerical example.
One of the main objectives of High Frequency Surface Wave Radar (HFSWR) is to adapt the spectrum of transmit waveform to avoid narrowband interferences of wireless communication systems. Sparse frequency waveform that contains several individual clear bands and stop bands is a desirable solution to this problem. Besides spectral requirements, good correlation properties of transmit waveform are also desired in specific applications, such as the range side lobe. Moreover, practical hardware constraints require real-time algorithm which has a fast convergence rate. In this paper, a new penalty function is proposed based on both requirements for the Power Spectrum Density (PSD) and the Integrated Sidelobe Level (ISL). Then, the phase-only conjugate gradient method with very fast convergence rate is proposed to solve this problem. The proposed approach is efficient in computation and flexible in designing waveform. Optimized waveform can effectively suppress narrowband interference and have low ISL. The validity of the proposed method is illustrated by numerical example.
2012, 34(6): 1297-1304.
doi: 10.3724/SP.J.1146.2011.00766
Abstract:
A fast method in polarimetric Synthetic Aperture Radar (SAR) data simulation for three-dimension forest application is proposed and the difficult application of SAR echo generation process in wide range and high density forest stand is solved. The method is based on the equivalent scattering process consisting of a few virtual scatters specially obtained by segmenting the entire forest stand into slices, blocks, and pieces. The received signal responded by a cluster of real scatters can be equivalent to that responded by a single virtual scatter in the case of the tiny slant range error. Then, the number of scatterers can be reduced and the efficiency of echo generation process can be improved. The error sensitivity of the equivalent parameters is analyzed and optimum combination of the parameters considering both precision and efficiency is obtained. The experimental results on 2.5105 m2 broadleaf forest stand show that the proposed method provides high efficiency simulation while keeping high precision on both echoes and images. The forest height inversion results on 1.0106 m2 broadleaf forest stand demonstrate the feasibility and potential utilities of the proposed method for large area forest applications.
A fast method in polarimetric Synthetic Aperture Radar (SAR) data simulation for three-dimension forest application is proposed and the difficult application of SAR echo generation process in wide range and high density forest stand is solved. The method is based on the equivalent scattering process consisting of a few virtual scatters specially obtained by segmenting the entire forest stand into slices, blocks, and pieces. The received signal responded by a cluster of real scatters can be equivalent to that responded by a single virtual scatter in the case of the tiny slant range error. Then, the number of scatterers can be reduced and the efficiency of echo generation process can be improved. The error sensitivity of the equivalent parameters is analyzed and optimum combination of the parameters considering both precision and efficiency is obtained. The experimental results on 2.5105 m2 broadleaf forest stand show that the proposed method provides high efficiency simulation while keeping high precision on both echoes and images. The forest height inversion results on 1.0106 m2 broadleaf forest stand demonstrate the feasibility and potential utilities of the proposed method for large area forest applications.
2012, 34(6): 1305-1310.
doi: 10.3724/SP.J.1146.2011.01121
Abstract:
In this paper, influence of azimuth non-uniform sampling on SAR systems with multiple-receive apertures in azimuth is derived theoretically. The analysis shows that several ghosting peaks are presented around the main peak, and at the same time non-uniform sampling will cause peak gain loss and the broadening to the main beam. Formulas to calculate the Peek-to-Ghost-Ratio (PGR) and peak gain loss are given, which show that they are related to the degree of non-uniform and the azimuth oversampling ratio quantitatively. The correctness of these formulas is validated further by simulation experiments.
In this paper, influence of azimuth non-uniform sampling on SAR systems with multiple-receive apertures in azimuth is derived theoretically. The analysis shows that several ghosting peaks are presented around the main peak, and at the same time non-uniform sampling will cause peak gain loss and the broadening to the main beam. Formulas to calculate the Peek-to-Ghost-Ratio (PGR) and peak gain loss are given, which show that they are related to the degree of non-uniform and the azimuth oversampling ratio quantitatively. The correctness of these formulas is validated further by simulation experiments.
2012, 34(6): 1311-1317.
doi: 10.3724/SP.J.1146.2011.00739
Abstract:
Airborne cross-track sparse array radar system for downward-looking three-dimensional (3D) imaging is investigated. The sparse array is distributed along the cross-track direction. To improve the transmitting power, orthogonal frequency division signals are used for Multiple-Input and Multiple-Output (MIMO). The multiple phase center aperture synthesis principle is used to make the sparse array coincident with a full array. The phase centers of different sub-band signals will be in the same position by using the matched filter related with spatial location. The broadband signal is obtained by combining the sub-band signals to improve the range resolution. And the downward-looking 3D image is achieved. To expand the observation swaths, a kind of scanning mode which combines ScanSAR mode and SweepSAR mode is proposed, which can reduce the Pulse Repetition Frequency (PRF). System parameters of the airborne cross-track sparse array radar are analyzed. The simulation results indicate the feasibility of the proposed method.
Airborne cross-track sparse array radar system for downward-looking three-dimensional (3D) imaging is investigated. The sparse array is distributed along the cross-track direction. To improve the transmitting power, orthogonal frequency division signals are used for Multiple-Input and Multiple-Output (MIMO). The multiple phase center aperture synthesis principle is used to make the sparse array coincident with a full array. The phase centers of different sub-band signals will be in the same position by using the matched filter related with spatial location. The broadband signal is obtained by combining the sub-band signals to improve the range resolution. And the downward-looking 3D image is achieved. To expand the observation swaths, a kind of scanning mode which combines ScanSAR mode and SweepSAR mode is proposed, which can reduce the Pulse Repetition Frequency (PRF). System parameters of the airborne cross-track sparse array radar are analyzed. The simulation results indicate the feasibility of the proposed method.
2012, 34(6): 1318-1323.
doi: 10.3724/SP.J.1146.2011.01094
Abstract:
For the problem of normal integrator exists serious collapsing loss when detecting spare scatterer targets, and Scatterer Density Depend-Binary Integrator (SDD-BI) exists signal to clutter ratio loss when detecting dense scatterer targets, a Double Threshold-Constant False Alarm Ratio (DT-CFAR) detector of range-spread target is proposed by using the density and amplitude of a target scatterer. The formula relating false alarm probability and detection probability to detection threshold is deduced in the Gaussian clutter background, and the optimal first threshold for different target parameters is given. Experiment results show that DT-CFAR has better detection performance for diverse scatterer density targets and has robustness when the scatter density of a target is estimated mismatching.
For the problem of normal integrator exists serious collapsing loss when detecting spare scatterer targets, and Scatterer Density Depend-Binary Integrator (SDD-BI) exists signal to clutter ratio loss when detecting dense scatterer targets, a Double Threshold-Constant False Alarm Ratio (DT-CFAR) detector of range-spread target is proposed by using the density and amplitude of a target scatterer. The formula relating false alarm probability and detection probability to detection threshold is deduced in the Gaussian clutter background, and the optimal first threshold for different target parameters is given. Experiment results show that DT-CFAR has better detection performance for diverse scatterer density targets and has robustness when the scatter density of a target is estimated mismatching.
2012, 34(6): 1324-1330.
doi: 10.3724/SP.J.1146.2011.00723
Abstract:
Channel blind equalization algorithm based on Eigen-Decomposition (ED) of data covariance matrix can be applied to calibrate the channel mismatch for multi-channel SAR/GMTI system. However, this algorithm has a disadvantage that it suffers from a slow convergence rate. In this paper, the principle of channel blind equalization algorithm is firstly investigated. Then, to improve its convergence rate, reduced-dimension technique is used into this algorithm and a new channel blind equalization algorithm is proposed. Experimental results on simulation data and measured SAR data demonstrate that compared with the conventional channel blind equalization algorithm, the proposed algorithm shows a fast convergence rate and is able to calibrate channel mismatch with much less sample support.
Channel blind equalization algorithm based on Eigen-Decomposition (ED) of data covariance matrix can be applied to calibrate the channel mismatch for multi-channel SAR/GMTI system. However, this algorithm has a disadvantage that it suffers from a slow convergence rate. In this paper, the principle of channel blind equalization algorithm is firstly investigated. Then, to improve its convergence rate, reduced-dimension technique is used into this algorithm and a new channel blind equalization algorithm is proposed. Experimental results on simulation data and measured SAR data demonstrate that compared with the conventional channel blind equalization algorithm, the proposed algorithm shows a fast convergence rate and is able to calibrate channel mismatch with much less sample support.
2012, 34(6): 1331-1336.
doi: 10.3724/SP.J.1146.2011.01074
Abstract:
The DPC-MAB SAR system enables the reconstruction of the azimuth signal with quite low PRF by coherent process of the multichannel signals, hence relieving the contradiction of the azimuth resolution and the swath for the traditional monochannel SAR. In this paper the multichannel reconstruction algorithms based on beamforming are investigated. Considering the disadvantages of the non-adaptive beamformer and the Capon beamformer, a novel Minimum Output Energy (MOE) beamforming algorithm is proposed, which can modify the space steering vectors adaptively, and can restrain the correlative ambiguities by the multiple restrictions. The effectiveness of this method is verified by the computer-based simulations and the imaging of the real SAR data.
The DPC-MAB SAR system enables the reconstruction of the azimuth signal with quite low PRF by coherent process of the multichannel signals, hence relieving the contradiction of the azimuth resolution and the swath for the traditional monochannel SAR. In this paper the multichannel reconstruction algorithms based on beamforming are investigated. Considering the disadvantages of the non-adaptive beamformer and the Capon beamformer, a novel Minimum Output Energy (MOE) beamforming algorithm is proposed, which can modify the space steering vectors adaptively, and can restrain the correlative ambiguities by the multiple restrictions. The effectiveness of this method is verified by the computer-based simulations and the imaging of the real SAR data.
2012, 34(6): 1337-1343.
doi: 10.3724/SP.J.1146.2011.01019
Abstract:
By stitching Doppler Beam Sharpening (DBS) images together, a wide swath ground image is obtained. In practice, however, there are some difficulties in stitching due to the nonideal movement (variation of the velocity vector and the attitude) of the radar platform. To deal with these problems, a novel method is proposed, which can stitch the DBS images effectively. The Interial Navigation System (INS) information is employed to estimate the instantaneous position of the aircraft and the radar beam direction, both of which are then transformed to the original reference coordinate so that the corrected parameters of the image can be obtained. Finally, by using these parameters, the image stitiching is achieved. Processing results of measured data show that the proposed method not only compensates the motion error excellently, but also significantly outperforms the traditional algorithm in stitching performance.
By stitching Doppler Beam Sharpening (DBS) images together, a wide swath ground image is obtained. In practice, however, there are some difficulties in stitching due to the nonideal movement (variation of the velocity vector and the attitude) of the radar platform. To deal with these problems, a novel method is proposed, which can stitch the DBS images effectively. The Interial Navigation System (INS) information is employed to estimate the instantaneous position of the aircraft and the radar beam direction, both of which are then transformed to the original reference coordinate so that the corrected parameters of the image can be obtained. Finally, by using these parameters, the image stitiching is achieved. Processing results of measured data show that the proposed method not only compensates the motion error excellently, but also significantly outperforms the traditional algorithm in stitching performance.
2012, 34(6): 1344-1350.
doi: 10.3724/SP.J.1146.2011.01097
Abstract:
High computational complexity is a problem that radar imaging technique based on Compressed Sensing (CS) must overcome for practical applications. In the light of the sparsity of radar target reflectivity, this paper studies 2D joint compressive imaging methods based on modified Orthogonal Matching Pursuit (OMP) algorithms. The sparse representation model of stepped frequency radar echo is established and analyzed, according to the 2D separability of sparse dictionary and compressive measurement, an improved OMP algorithm is proposed for radar image formation, which improves the computational efficiency greatly and can be extended to other greedy algorithms easily. Theoretical comparison and analysis indicate that the proposed methods possess prominent superiority over storage and computation compared to conventional CS algorithms, experiments from both simulated data and measured data verify their validity.
High computational complexity is a problem that radar imaging technique based on Compressed Sensing (CS) must overcome for practical applications. In the light of the sparsity of radar target reflectivity, this paper studies 2D joint compressive imaging methods based on modified Orthogonal Matching Pursuit (OMP) algorithms. The sparse representation model of stepped frequency radar echo is established and analyzed, according to the 2D separability of sparse dictionary and compressive measurement, an improved OMP algorithm is proposed for radar image formation, which improves the computational efficiency greatly and can be extended to other greedy algorithms easily. Theoretical comparison and analysis indicate that the proposed methods possess prominent superiority over storage and computation compared to conventional CS algorithms, experiments from both simulated data and measured data verify their validity.
2012, 34(6): 1351-1355.
doi: 10.3724/SP.J.1146.2011.00954
Abstract:
The Circular Synthetic Aperture Radar (CSAR), unlike the linear SAR, has the capability to extract three-dimensional imaging information of a target scene. In this paper, a novel CSAR wavefront reconstruction method is proposed for three-dimensional imaging. The proposed method compensates the effects of the phase components introduced by the radars trajectory to form reconstructed image, which avoids calculating the complicated Hankel function. Thus the complexity of the algorithm can be reduced. Finally, the validity of the presented method is clarified by both simulated data and experimental data sets.
The Circular Synthetic Aperture Radar (CSAR), unlike the linear SAR, has the capability to extract three-dimensional imaging information of a target scene. In this paper, a novel CSAR wavefront reconstruction method is proposed for three-dimensional imaging. The proposed method compensates the effects of the phase components introduced by the radars trajectory to form reconstructed image, which avoids calculating the complicated Hankel function. Thus the complexity of the algorithm can be reduced. Finally, the validity of the presented method is clarified by both simulated data and experimental data sets.
2012, 34(6): 1356-1361.
doi: 10.3724/SP.J.1146.2011.01129
Abstract:
The similarity measures between SAR images are the basis for target recognition, classification, image registration and so on. The interpretation of SAR images can be improved obviously by defining reasonable and reliably similarity measure. In this paper, a confidence interval of similarity and the method of constructing credibility are proposed based on the characteristics of image contours. Firstly, the similarity measure is obtained by fuzzifying the uncertain contours in SAR images. Then, the confidence interval of the measure as well as the definition of credibility at a given confidence level are got by analyzing the distribution function of fuzzy model. The experimental results demonstrate that the proposed similarity measure between a given pair of images is robust to the location of image contours. Besides, it can also provide reasonable similarity with credibility of the fractured and multi-edged contours, which has not been presented by other methods. Moreover, it is consistent with human visual perception.
The similarity measures between SAR images are the basis for target recognition, classification, image registration and so on. The interpretation of SAR images can be improved obviously by defining reasonable and reliably similarity measure. In this paper, a confidence interval of similarity and the method of constructing credibility are proposed based on the characteristics of image contours. Firstly, the similarity measure is obtained by fuzzifying the uncertain contours in SAR images. Then, the confidence interval of the measure as well as the definition of credibility at a given confidence level are got by analyzing the distribution function of fuzzy model. The experimental results demonstrate that the proposed similarity measure between a given pair of images is robust to the location of image contours. Besides, it can also provide reasonable similarity with credibility of the fractured and multi-edged contours, which has not been presented by other methods. Moreover, it is consistent with human visual perception.
2012, 34(6): 1362-1367.
doi: 10.3724/SP.J.1146.2011.01328
Abstract:
In this paper, an improved rank-reduction algorithm based on Householder Multi-stage Weiner Filter (Householder-MWF) is proposed for the application of space-time anti-jam in GPS receivers. The new algorithm retains the forward decomposition process of the original algorithm and improves its backward recursion by utilizing the decomposition property of Multi-stage Weiner Filtering (MWF), so obtains an order-recursive implementation structure and greatly promotes the real-time performance. Its shown that the computation complexity of the proposed algorithm is almost the same as the original algorithm and is apparently lower than algorithms based on the Correlation Subtraction Structure (CSS) MWF. The anti-jam performance of the proposed algorithm is verified by simulations.
In this paper, an improved rank-reduction algorithm based on Householder Multi-stage Weiner Filter (Householder-MWF) is proposed for the application of space-time anti-jam in GPS receivers. The new algorithm retains the forward decomposition process of the original algorithm and improves its backward recursion by utilizing the decomposition property of Multi-stage Weiner Filtering (MWF), so obtains an order-recursive implementation structure and greatly promotes the real-time performance. Its shown that the computation complexity of the proposed algorithm is almost the same as the original algorithm and is apparently lower than algorithms based on the Correlation Subtraction Structure (CSS) MWF. The anti-jam performance of the proposed algorithm is verified by simulations.
2012, 34(6): 1368-1373.
doi: 10.3724/SP.J.1146.2011.00941
Abstract:
Multi-target filtering using Probability Hypothesis Density (PHD) in multi-sensor case is based on assumption model to avoid being computationally intractable. Based on describing target state space and sensor observation space by Random Finite Set (RFS) method, and on the analysis of detection probability, likelihood function and clutter distribution under the multi-sensor universal assumption model, the multi-sensor version of multi-target PHD filter is constructed by Probability Generating Functional (PGFL), the multi-sensor labeling particle Sequential Monte Carlo PHD (SMC-PHD) filtering algorithm is presented to implement this fiter with lower computational complexity. Finally, the better estimation of target number and track-valued state are obtained by simulation.
Multi-target filtering using Probability Hypothesis Density (PHD) in multi-sensor case is based on assumption model to avoid being computationally intractable. Based on describing target state space and sensor observation space by Random Finite Set (RFS) method, and on the analysis of detection probability, likelihood function and clutter distribution under the multi-sensor universal assumption model, the multi-sensor version of multi-target PHD filter is constructed by Probability Generating Functional (PGFL), the multi-sensor labeling particle Sequential Monte Carlo PHD (SMC-PHD) filtering algorithm is presented to implement this fiter with lower computational complexity. Finally, the better estimation of target number and track-valued state are obtained by simulation.
2012, 34(6): 1374-1379.
doi: 10.3724/SP.J.1146.2011.00893
Abstract:
Compressive imaging is a novel imaging method based on compressive sensing theory, the key idea is that it can reconstruct original scene precisely with far fewer measurements than Nyquist samples if the scene is sparse/compressible; Constructing an appropriate measurement matrix easy to realize random linear measurement of an image is one of the key points of practical compressive sensing. In this paper, analyzing the existing Bernoulli and Circulant matrices, a novel sparse trinary circulant measurement matrix with random spacing for phase mask is proposed. Simulation results show that novel phase mask matrices, compared to Bernoulli and Bernoulli-Circulant (BC) phase mask matrices, have the same signal-to-noise ratio; But with the number of independent random variables and the number of non-zeros entries a dramatically reduction, which is more conducive to data transmission and storage; more importantly that is easy to hardware implementation and the reconstructed time is only about 20%~50% of that of original matrices, which has a significance effects on practical compressive sensing.
Compressive imaging is a novel imaging method based on compressive sensing theory, the key idea is that it can reconstruct original scene precisely with far fewer measurements than Nyquist samples if the scene is sparse/compressible; Constructing an appropriate measurement matrix easy to realize random linear measurement of an image is one of the key points of practical compressive sensing. In this paper, analyzing the existing Bernoulli and Circulant matrices, a novel sparse trinary circulant measurement matrix with random spacing for phase mask is proposed. Simulation results show that novel phase mask matrices, compared to Bernoulli and Bernoulli-Circulant (BC) phase mask matrices, have the same signal-to-noise ratio; But with the number of independent random variables and the number of non-zeros entries a dramatically reduction, which is more conducive to data transmission and storage; more importantly that is easy to hardware implementation and the reconstructed time is only about 20%~50% of that of original matrices, which has a significance effects on practical compressive sensing.
2012, 34(6): 1380-1387.
doi: 10.3724/SP.J.1146.2011.00932
Abstract:
Recently, it is a new direction to improve the performance of image steganalysis by combining the detection of information hidden with image content analysis. Relative to methods depending on entire image, this paper analyzes the effect of LSB (Least Significant Bit) matching steganography on image sub-areas, and presents a novel steganalyzer based on the combined discrimination. Firstly, the images are divided into several sub-areas according to the image content complexity. Secondly, the histogram spectral features of pixel sequence of each sub-area are extracted by using two different filters. Then, the Bayes classifiers are trained respectively by features of each class of sub-area in order to obtain its weights. Finally, each sub-area of a test image is detected depending on its class and the final discrimination result of the test image is achieved by weighted fusion of the results of its sub-areas. Experimental results show that the proposed method exhibits excellent performance for the detection of LSB matching, outperforms existing representative approaches.
Recently, it is a new direction to improve the performance of image steganalysis by combining the detection of information hidden with image content analysis. Relative to methods depending on entire image, this paper analyzes the effect of LSB (Least Significant Bit) matching steganography on image sub-areas, and presents a novel steganalyzer based on the combined discrimination. Firstly, the images are divided into several sub-areas according to the image content complexity. Secondly, the histogram spectral features of pixel sequence of each sub-area are extracted by using two different filters. Then, the Bayes classifiers are trained respectively by features of each class of sub-area in order to obtain its weights. Finally, each sub-area of a test image is detected depending on its class and the final discrimination result of the test image is achieved by weighted fusion of the results of its sub-areas. Experimental results show that the proposed method exhibits excellent performance for the detection of LSB matching, outperforms existing representative approaches.
2012, 34(6): 1388-1392.
doi: 10.3724/SP.J.1146.2011.00942
Abstract:
Fractional calculus is widely used in control system theory. Owing to introducing of fractional dynamic system model, searching solution method of fractional state estimation is an urgent issue. Starting from fractional nonlinear dynamic system model, fractional unscented Kalman filter is derived based on probability theory. The filter is applied to two typical nonlinear systems, Univariate Nonstationary Growth Model (UNGM) model and Reentry Vehicle Tracking (RVT) model. Experiment results prove the performance of fractional unscented Kalman filter given in this paper is better than that of the unscented Kalman filter in the context of reasonable setting of the fractional order.
Fractional calculus is widely used in control system theory. Owing to introducing of fractional dynamic system model, searching solution method of fractional state estimation is an urgent issue. Starting from fractional nonlinear dynamic system model, fractional unscented Kalman filter is derived based on probability theory. The filter is applied to two typical nonlinear systems, Univariate Nonstationary Growth Model (UNGM) model and Reentry Vehicle Tracking (RVT) model. Experiment results prove the performance of fractional unscented Kalman filter given in this paper is better than that of the unscented Kalman filter in the context of reasonable setting of the fractional order.
2012, 34(6): 1393-1397.
doi: 10.3724/SP.J.1146.2011.01039
Abstract:
In order to achieve the optimal detection algorithm for interference alignment, a diversity detection algorithm is proposed based on analyzing the feature of signal subspace. In this algorithm, the projection of desired signal onto signal subspace is served as detection vector to maximize received signal-to-noise-ratio. Numerical results show that the proposed algorithm outperforms existing schemes when the interference alignment conditions are satisfied strictly. Furthermore, with the increasing of diversity degree, the benefit of the algorithm becomes more and more obvious.
In order to achieve the optimal detection algorithm for interference alignment, a diversity detection algorithm is proposed based on analyzing the feature of signal subspace. In this algorithm, the projection of desired signal onto signal subspace is served as detection vector to maximize received signal-to-noise-ratio. Numerical results show that the proposed algorithm outperforms existing schemes when the interference alignment conditions are satisfied strictly. Furthermore, with the increasing of diversity degree, the benefit of the algorithm becomes more and more obvious.
2012, 34(6): 1398-1403.
doi: 10.3724/SP.J.1146.2011.00683
Abstract:
In the parallelization of Space-Time Adaptive Processing (STAP) arithmetic, traditional methods schedule the STAP arithmetic to different processors in the specific hardware architecture through coral-granularity division and improve the throughput by pipeline processing between processors. In the paper, its disadvantages are discussed from two perspectives: Coarse-grained scheduling hinders the parallelism; They are only suitable for the specific system parameters and hardware architectures. Thus, a new method based on fine-grained scheduling is put forward, which consists of three steps: Firstly, fine-grained task model in the form of Direct Acyclic Graph (DAG) is constructed; Secondly, the topology model is built to describe the target system; Finally, the established task model in fine-grained manner is assigned to different processors described in model topology. The experiment of the proposed method shows that it achieves better acceleration ratio, and more flexiable adaptation to different STAP applications.
In the parallelization of Space-Time Adaptive Processing (STAP) arithmetic, traditional methods schedule the STAP arithmetic to different processors in the specific hardware architecture through coral-granularity division and improve the throughput by pipeline processing between processors. In the paper, its disadvantages are discussed from two perspectives: Coarse-grained scheduling hinders the parallelism; They are only suitable for the specific system parameters and hardware architectures. Thus, a new method based on fine-grained scheduling is put forward, which consists of three steps: Firstly, fine-grained task model in the form of Direct Acyclic Graph (DAG) is constructed; Secondly, the topology model is built to describe the target system; Finally, the established task model in fine-grained manner is assigned to different processors described in model topology. The experiment of the proposed method shows that it achieves better acceleration ratio, and more flexiable adaptation to different STAP applications.
2012, 34(6): 1404-1407.
doi: 10.3724/SP.J.1146.2011.01139
Abstract:
An algorithm of speaker clustering is proposed based on Feature Mean Distance (FMD) for short speech segments. First, a distance measure, i.e. FMD, is introduced to represent the similarities between two clusters on the level of feature instead of the level of model. Then, two clusters with the minimum of FMDs are iteratively merged until the minimum of FMDs is larger than an adaptive threshold. Experimental results show average 5% improvements in F measure are obtained in comparison with the AHC+BIC based algorithm. In addition, the proposed algorithm is 4.68 times faster than the AHC+BIC based algorithm.
An algorithm of speaker clustering is proposed based on Feature Mean Distance (FMD) for short speech segments. First, a distance measure, i.e. FMD, is introduced to represent the similarities between two clusters on the level of feature instead of the level of model. Then, two clusters with the minimum of FMDs are iteratively merged until the minimum of FMDs is larger than an adaptive threshold. Experimental results show average 5% improvements in F measure are obtained in comparison with the AHC+BIC based algorithm. In addition, the proposed algorithm is 4.68 times faster than the AHC+BIC based algorithm.
2012, 34(6): 1408-1413.
doi: 10.3724/SP.J.1146.2011.01001
Abstract:
On the research of compressed sensing, the sparse field by certain transformations is one of the most important factors on signal reconstruction. This paper presents a new united sparsity method based on Linear Prediction Coefficients (LPC) of speech signals, which associates LPC analysis with difference transform method. Orthogonal Matching Pursuit (OMP) algorithm is used to reconstruct the speech signal, and the reconstruction performance by this new method is compared with FFT and LPC. Experiments show that, when the compression ratio is larger than 0.4, the performance of reconstructed signal by united method is much better than the other two. Namely, when the reconstruction performance of the three methods is same, the compression ratio of the united method is less than that of the two, which means the united method has better compression performance. PESQ is used to evaluate the quality of reconstructed speech, and the speech reconstructed by the united method has the higher scores.
On the research of compressed sensing, the sparse field by certain transformations is one of the most important factors on signal reconstruction. This paper presents a new united sparsity method based on Linear Prediction Coefficients (LPC) of speech signals, which associates LPC analysis with difference transform method. Orthogonal Matching Pursuit (OMP) algorithm is used to reconstruct the speech signal, and the reconstruction performance by this new method is compared with FFT and LPC. Experiments show that, when the compression ratio is larger than 0.4, the performance of reconstructed signal by united method is much better than the other two. Namely, when the reconstruction performance of the three methods is same, the compression ratio of the united method is less than that of the two, which means the united method has better compression performance. PESQ is used to evaluate the quality of reconstructed speech, and the speech reconstructed by the united method has the higher scores.
2012, 34(6): 1414-1419.
doi: 10.3724/SP.J.1146.2011.01045
Abstract:
An - insensitive criterion and structure risk based Radius-Basis-Function Neural-Network (RBF-NN) modeling method is proposed. By -introducing insensitive criterion and the item of structure risk, the RBF-NN learning is transformed into the linear regression and Quadratic Program (QP) optimization issue. Compared with the traditional least-square-criterion based RBF-NN training algorithms, the proposed method is much more robust to noise data and small size of datasets. Through the simulation experiments on the synthetic and real-word datasets, the above virtues are confirmed.
An - insensitive criterion and structure risk based Radius-Basis-Function Neural-Network (RBF-NN) modeling method is proposed. By -introducing insensitive criterion and the item of structure risk, the RBF-NN learning is transformed into the linear regression and Quadratic Program (QP) optimization issue. Compared with the traditional least-square-criterion based RBF-NN training algorithms, the proposed method is much more robust to noise data and small size of datasets. Through the simulation experiments on the synthetic and real-word datasets, the above virtues are confirmed.
2012, 34(6): 1420-1424.
doi: 10.3724/SP.J.1146.2011.00929
Abstract:
A game-theoretic analysis is performed to acquire the equilibrium strategies and their precondition by formalizing the strategies and payoffs of players in the honeypot game. The effectiveness and the deficiency of honeypot are inferred for network confrontation game, which demonstrate that honeypot is a passive-proactive defense mechanism. The ideal equilibrium combined strategy for defender and the factor affecting the effectiveness and activeness are discussed in detail. Inspired by the mimicry phenomena in biological competition, the propositional methods are given to enhance the deception performance for honeypot. The work is a theoretical support for the effective and proactive tactics of honeypot.
A game-theoretic analysis is performed to acquire the equilibrium strategies and their precondition by formalizing the strategies and payoffs of players in the honeypot game. The effectiveness and the deficiency of honeypot are inferred for network confrontation game, which demonstrate that honeypot is a passive-proactive defense mechanism. The ideal equilibrium combined strategy for defender and the factor affecting the effectiveness and activeness are discussed in detail. Inspired by the mimicry phenomena in biological competition, the propositional methods are given to enhance the deception performance for honeypot. The work is a theoretical support for the effective and proactive tactics of honeypot.
2012, 34(6): 1425-1431.
doi: 10.3724/SP.J.1146.2011.01052
Abstract:
To address the issue that most link loss rate inference algorithm can not take into account the accuracy and speed, a novel link loss rate inference algorithm of overlay network is proposed. This algorithm converts the link loss rate inference issue into solution of nonlinear programming issue in order to calculate the packet loss rate of all links more accurately and only needs to probe once for every path, with no extra deployment costs. Through software simulations and Internet experiments, the feasibility and superiority of this algorithm is proved.
To address the issue that most link loss rate inference algorithm can not take into account the accuracy and speed, a novel link loss rate inference algorithm of overlay network is proposed. This algorithm converts the link loss rate inference issue into solution of nonlinear programming issue in order to calculate the packet loss rate of all links more accurately and only needs to probe once for every path, with no extra deployment costs. Through software simulations and Internet experiments, the feasibility and superiority of this algorithm is proved.
2012, 34(6): 1432-1437.
doi: 10.3724/SP.J.1146.2011.01170
Abstract:
The relations among objects with secure level are very complex, which leads to the problems of security in multi-level network, such as information leakage by object aggregation. This paper puts forward a level inference method for aggregated information of objects based on clustering analysis. This method makes the reduction of attributes by the importance degree of attribute in one object, and attribute vector is formed. Then, according to formal concept analysis, this method accomplishes comparability analysis of objects in the same secure domain by gravity among concepts so that objects can be aggregated. Finally, according to probability estimate of fuzzy set about secure level of attributes or sets of attribute, probability of higher level information inferred by aggregation of similar objects is computed, which may effectively establish access control policy in multi-level network, and accomplish restricted access of congeneric objects in order to reduce the risk of information system.
The relations among objects with secure level are very complex, which leads to the problems of security in multi-level network, such as information leakage by object aggregation. This paper puts forward a level inference method for aggregated information of objects based on clustering analysis. This method makes the reduction of attributes by the importance degree of attribute in one object, and attribute vector is formed. Then, according to formal concept analysis, this method accomplishes comparability analysis of objects in the same secure domain by gravity among concepts so that objects can be aggregated. Finally, according to probability estimate of fuzzy set about secure level of attributes or sets of attribute, probability of higher level information inferred by aggregation of similar objects is computed, which may effectively establish access control policy in multi-level network, and accomplish restricted access of congeneric objects in order to reduce the risk of information system.
2012, 34(6): 1438-1444.
doi: 10.3724/SP.J.1146.2011.01084
Abstract:
In order to reflect the actual channel attenuation, a model selection algorithm for path loss in wireless sensor networks is proposed. Firstly, the statistical properties of some path loss models are analyzed,and then Expectation Maximization (EM) algorithm from incomplete data of Received Signal Strength (RSS) is proposed for parameter estimation. Finally, a set of weighted coefficients are given on the basis of criterion function, which could select an appropriate path loss model. Through experiment, the proposed model selection method could estimate parameters effectively, compared with other similar algorithms,this method could pick up a model fitting the experimental data better.
In order to reflect the actual channel attenuation, a model selection algorithm for path loss in wireless sensor networks is proposed. Firstly, the statistical properties of some path loss models are analyzed,and then Expectation Maximization (EM) algorithm from incomplete data of Received Signal Strength (RSS) is proposed for parameter estimation. Finally, a set of weighted coefficients are given on the basis of criterion function, which could select an appropriate path loss model. Through experiment, the proposed model selection method could estimate parameters effectively, compared with other similar algorithms,this method could pick up a model fitting the experimental data better.
2012, 34(6): 1445-1451.
doi: 10.3724/SP.J.1146.2011.01060
Abstract:
In order to estimate the software trustworthiness accurately, a trustworthy behavior model for software monitoring point based on classification attributes is proposed for the software monitoring point in the expected behavior trace of software. Firstly, the attributes of software monitoring point are classified according to the sphere of action during the trustworthiness evaluation and the trustworthy behavior model of each attribute level is constructed. Secondly, for scene level attributes, a clustering algorithm of scene level attributes based on Gaussian kernel function is presented considering the distinction of training samples of one monitoring point, and a weight distribution strategy for scene level attributes based on one-class samples is proposed for one-class training samples. Finally, experiments and analyses show that: the model can evaluate software monitoring point accurately; For trustworthy behavior model of scene level attributes, the clustering algorithm has lower classification error rate than other clustering algorithms, and the weight distribution strategy has better effect of trustworthiness evaluation than other methods of weight distribution for one-class samples.
In order to estimate the software trustworthiness accurately, a trustworthy behavior model for software monitoring point based on classification attributes is proposed for the software monitoring point in the expected behavior trace of software. Firstly, the attributes of software monitoring point are classified according to the sphere of action during the trustworthiness evaluation and the trustworthy behavior model of each attribute level is constructed. Secondly, for scene level attributes, a clustering algorithm of scene level attributes based on Gaussian kernel function is presented considering the distinction of training samples of one monitoring point, and a weight distribution strategy for scene level attributes based on one-class samples is proposed for one-class training samples. Finally, experiments and analyses show that: the model can evaluate software monitoring point accurately; For trustworthy behavior model of scene level attributes, the clustering algorithm has lower classification error rate than other clustering algorithms, and the weight distribution strategy has better effect of trustworthiness evaluation than other methods of weight distribution for one-class samples.
2012, 34(6): 1452-1456.
doi: 10.3724/SP.J.1146.2011.01235
Abstract:
The theory of compressed sensing can be used to reconstruct sparse signals from fewer observations. According to the sparsity of UWB channels, a reduced sampling rate can be obtained at the detector based on compressed sensing frame. In this paper, a filter matrix estimation algorithm is proposed by designing the over-completed dictionary and observation matrix. Then, the Orthogonal Matching Pursuit (OMP), the Basis Pursuit De-noising (BPDN) and the Dantzig Selector (DS) are used to detect original signal to give the opinions for choosing suitable reconstruction algorithms. The simulation results in the IEEE 802.15.4a channel model show that the coherence detection based on the new channel estimation method outperforms the one based on random observation method for better bit error rate performances with a reduced sampling rate.
The theory of compressed sensing can be used to reconstruct sparse signals from fewer observations. According to the sparsity of UWB channels, a reduced sampling rate can be obtained at the detector based on compressed sensing frame. In this paper, a filter matrix estimation algorithm is proposed by designing the over-completed dictionary and observation matrix. Then, the Orthogonal Matching Pursuit (OMP), the Basis Pursuit De-noising (BPDN) and the Dantzig Selector (DS) are used to detect original signal to give the opinions for choosing suitable reconstruction algorithms. The simulation results in the IEEE 802.15.4a channel model show that the coherence detection based on the new channel estimation method outperforms the one based on random observation method for better bit error rate performances with a reduced sampling rate.
2012, 34(6): 1457-1462.
doi: 10.3724/SP.J.1146.2011.01237
Abstract:
The intrinsic characteristics of wireless Ad hoc network, such as self-organizing, self-configuration and self-adaptation, make this kind of network an attractive solution in several scenarios in which lack of a fixed infrastructure. How to evaluate the capacity performance becomes one of the hop topics in the research area of wireless Ad hoc networks. In this paper, stochastic geometry and Poisson point process are applied to establish the network model under different medium access control schemes, namely un-slotted, slotted ALOHA, and CSMA. Based on different modulation/coding schemes and assuming channel inversion power control strategy, outage probability and transmission capacity of networks are investigated. The numerical results show that under a constraint of maximum packet error probability, the choices of the modulation and coding schemes that maximize the transmission capacity, depend on the density of transmissions, and that is, the interference level of the network.
The intrinsic characteristics of wireless Ad hoc network, such as self-organizing, self-configuration and self-adaptation, make this kind of network an attractive solution in several scenarios in which lack of a fixed infrastructure. How to evaluate the capacity performance becomes one of the hop topics in the research area of wireless Ad hoc networks. In this paper, stochastic geometry and Poisson point process are applied to establish the network model under different medium access control schemes, namely un-slotted, slotted ALOHA, and CSMA. Based on different modulation/coding schemes and assuming channel inversion power control strategy, outage probability and transmission capacity of networks are investigated. The numerical results show that under a constraint of maximum packet error probability, the choices of the modulation and coding schemes that maximize the transmission capacity, depend on the density of transmissions, and that is, the interference level of the network.
2012, 34(6): 1463-1468.
doi: 10.3724/SP.J.1146.2011.01073
Abstract:
In spectrum sharing cognitive relay networks, due to the presence of imperfect Channel State Information (CSI), the interference from the secondary user can be higher than the maximum allowable interference level at the primary user and affect the primary users reliable communication. In order to quantify the impact of the secondary users interference on the primary user, a performance metric termed as interference probability is proposed. The closed-form expressions for interference probability of Opportunistic Relaying (OR) and Selection Cooperation (SC) are derived with imperfect CSI. It is shown that, in terms of the interference probability, the SC is better than the OR. However, the limit interference probability of the OR is identical to that of the SC. Moreover, with imperfect CSI, for both the SC and the OR, increasing the number of relays means a higher interference level on the primary user. Finally, simulation results are presented to verify the analysis.
In spectrum sharing cognitive relay networks, due to the presence of imperfect Channel State Information (CSI), the interference from the secondary user can be higher than the maximum allowable interference level at the primary user and affect the primary users reliable communication. In order to quantify the impact of the secondary users interference on the primary user, a performance metric termed as interference probability is proposed. The closed-form expressions for interference probability of Opportunistic Relaying (OR) and Selection Cooperation (SC) are derived with imperfect CSI. It is shown that, in terms of the interference probability, the SC is better than the OR. However, the limit interference probability of the OR is identical to that of the SC. Moreover, with imperfect CSI, for both the SC and the OR, increasing the number of relays means a higher interference level on the primary user. Finally, simulation results are presented to verify the analysis.
2012, 34(6): 1469-1475.
doi: 10.3724/SP.J.1146.2011.00674
Abstract:
As the existing clustering algorithms are lack of effective guidable clustering model, an Degree of Wiliness to Cooperate (DWC) based clustering model is proposed, in which the clustering objective of maximizing the system sum rate is approximately to maximizing the sum of DWC between every two Base Stations (BS) in system. Based on this, the clustering issue is modeled as constructing benefit-trees of a connected graph with edge costs. Then a benefit-tree dynamic clustering algorithm is proposed. This algorithm simultaneously generates several clusters of dynamic size which could solve the limited-capacity problem caused by conventional orderly clustering scheme. Besides, the maximum sum of DWC in clustering results offers the approximately best system clustering capacity. Simulation results show that compared to the conventional greedy clustering algorithm, the system spectrum efficiency in this algorithm increases about 0.4 bit/Hz and the computational complexity is directly proportional to system size.
As the existing clustering algorithms are lack of effective guidable clustering model, an Degree of Wiliness to Cooperate (DWC) based clustering model is proposed, in which the clustering objective of maximizing the system sum rate is approximately to maximizing the sum of DWC between every two Base Stations (BS) in system. Based on this, the clustering issue is modeled as constructing benefit-trees of a connected graph with edge costs. Then a benefit-tree dynamic clustering algorithm is proposed. This algorithm simultaneously generates several clusters of dynamic size which could solve the limited-capacity problem caused by conventional orderly clustering scheme. Besides, the maximum sum of DWC in clustering results offers the approximately best system clustering capacity. Simulation results show that compared to the conventional greedy clustering algorithm, the system spectrum efficiency in this algorithm increases about 0.4 bit/Hz and the computational complexity is directly proportional to system size.
2012, 34(6): 1476-1482.
doi: 10.3724/SP.J.1146.2011.01187
Abstract:
Based on the general element active pattern achieved by full wave simulation, a novel Adaptive Dynamic Meta Particle Swarm Optimization (ADMPSO) algorithm is proposed for synthesizing multiple-pattern conformal array. The dominated subgroup and nondominated subgroup are defined on the basis of traditional Meta particle swarm. Meanwhile, the adaptive dynamic modulating for multiple-subgroup is realized by introducing the downsizing of nondominated subgroup, the dominated subgroup expansion, and the updating for the adaptive inertia weights, which improve the optimization performance of ADMPSO. Furthermore, ADMPSO algorithm is applied to synthesizing multiple-pattern in 12-element microstrip conic sector conformal array in non-equatorial plane successfully, with the polarizing deterioration of the element considered. And the pencil beam, the flat-top beam, and the cosecant squared beam patterns are achieved simultaneously, with the common amplitude of the excitation restricted. The simulation shows that the optimization results and convergent speed are significantly improved compared with other algorithms.
Based on the general element active pattern achieved by full wave simulation, a novel Adaptive Dynamic Meta Particle Swarm Optimization (ADMPSO) algorithm is proposed for synthesizing multiple-pattern conformal array. The dominated subgroup and nondominated subgroup are defined on the basis of traditional Meta particle swarm. Meanwhile, the adaptive dynamic modulating for multiple-subgroup is realized by introducing the downsizing of nondominated subgroup, the dominated subgroup expansion, and the updating for the adaptive inertia weights, which improve the optimization performance of ADMPSO. Furthermore, ADMPSO algorithm is applied to synthesizing multiple-pattern in 12-element microstrip conic sector conformal array in non-equatorial plane successfully, with the polarizing deterioration of the element considered. And the pencil beam, the flat-top beam, and the cosecant squared beam patterns are achieved simultaneously, with the common amplitude of the excitation restricted. The simulation shows that the optimization results and convergent speed are significantly improved compared with other algorithms.
2012, 34(6): 1489-1493.
doi: 10.3724/SP.J.1146.2011.01151
Abstract:
This paper presents two different dual-band cavity filters based on the frequency transformation. The cavity filters have two different configurations. The one is presented by directed coupling between adjacent resonators, and the other one is presented by introduced cross coupling between nonadjacent resonators to generate a pair of transmission zeros. At first, according to the filter specification the coupling coefficients between resonators and the external quality factors are calculated and obtained. Then the model of the dual-band cavity filters are established with Ansoft HFSS, the simulation result agrees well with the result of equal circuit in ADS, indicating that this method is correct and feasible. Finally, the comparison between the cascade type structure and cross coupling type structure shows that the later has more excellent stopband performance.
This paper presents two different dual-band cavity filters based on the frequency transformation. The cavity filters have two different configurations. The one is presented by directed coupling between adjacent resonators, and the other one is presented by introduced cross coupling between nonadjacent resonators to generate a pair of transmission zeros. At first, according to the filter specification the coupling coefficients between resonators and the external quality factors are calculated and obtained. Then the model of the dual-band cavity filters are established with Ansoft HFSS, the simulation result agrees well with the result of equal circuit in ADS, indicating that this method is correct and feasible. Finally, the comparison between the cascade type structure and cross coupling type structure shows that the later has more excellent stopband performance.
2012, 34(6): 1494-1500.
doi: 10.3724/SP.J.1146.2011.00959
Abstract:
To solve the problem of scalability in designing polymorphic self-checking circuits using evolutionary design, a new method based on inputs decomposition and outputs matching is proposed. The system is decomposed into evolvable part and fixed part，and the number of input-output combinations can be decreased by decomposing the inputs of the system, thus the complexity of evolution is reduced. The NOT gate is added to outputs of the candidate circuits when the matching degree of the outputs is lower than 1/2 compared with desired outputs in the stage of fitness evaluation. The fitness as well as the diversity of the population is increased, and the optimum structure is protected from being eliminated. The evolutionary design experiments for two kind of self-checking adders are conducted by combining the polymorphic gates with ordinary gates. The results show that the generation of evolution is decreased by 47.9% and 89.1% while the fault coverage of single test vector is decreased by 75.7% and 79.7% compared with conventional method in designing polymorphic self-checking circuits. The proposed method enjoys advantages of faster convergence, better scalability and higher fault coverage.
To solve the problem of scalability in designing polymorphic self-checking circuits using evolutionary design, a new method based on inputs decomposition and outputs matching is proposed. The system is decomposed into evolvable part and fixed part，and the number of input-output combinations can be decreased by decomposing the inputs of the system, thus the complexity of evolution is reduced. The NOT gate is added to outputs of the candidate circuits when the matching degree of the outputs is lower than 1/2 compared with desired outputs in the stage of fitness evaluation. The fitness as well as the diversity of the population is increased, and the optimum structure is protected from being eliminated. The evolutionary design experiments for two kind of self-checking adders are conducted by combining the polymorphic gates with ordinary gates. The results show that the generation of evolution is decreased by 47.9% and 89.1% while the fault coverage of single test vector is decreased by 75.7% and 79.7% compared with conventional method in designing polymorphic self-checking circuits. The proposed method enjoys advantages of faster convergence, better scalability and higher fault coverage.
2012, 34(6): 1501-1505.
doi: 10.3724/SP.J.1146.2011.00905
Abstract:
A new POLSAR dominant scattering mechanism clustering and unsupervised classification method is proposed based on polarimetric scattering similarity and Freeman decomposition models. As polarimetric scattering similarity is roll-invariant and nonnegative, the method fixes the problems of Freeman decomposition in classification. Experiment shows that, comparing to the classification method based on Freeman decomposition, the proposed method depicts terrain scattering properties more accurately.
A new POLSAR dominant scattering mechanism clustering and unsupervised classification method is proposed based on polarimetric scattering similarity and Freeman decomposition models. As polarimetric scattering similarity is roll-invariant and nonnegative, the method fixes the problems of Freeman decomposition in classification. Experiment shows that, comparing to the classification method based on Freeman decomposition, the proposed method depicts terrain scattering properties more accurately.
2012, 34(6): 1506-1510.
doi: 10.3724/SP.J.1146.2011.01171
Abstract:
The direction of gradient vector is a factor being omitted in traditional seam carving. A new seam carving algorithm is proposed in this paper based on gradient vector direction analysis. A low-pass filter is firstly used on the local gradient vector to extract more suitable seams based on the fact that the gradient vector directions in random texture regions are usually disordered. Then two different pixel energy functions are defined, which give different weights to the x, y direction components of pixels gradient vectors based on different path direction of seams. The experiment results show that the method can better protect the image edge and other details and can reach more similar overall visual effect with original images. Quantitative analysis results also show that the algorithm achieves better results in both image completeness distance and image coherence distance than other methods.
The direction of gradient vector is a factor being omitted in traditional seam carving. A new seam carving algorithm is proposed in this paper based on gradient vector direction analysis. A low-pass filter is firstly used on the local gradient vector to extract more suitable seams based on the fact that the gradient vector directions in random texture regions are usually disordered. Then two different pixel energy functions are defined, which give different weights to the x, y direction components of pixels gradient vectors based on different path direction of seams. The experiment results show that the method can better protect the image edge and other details and can reach more similar overall visual effect with original images. Quantitative analysis results also show that the algorithm achieves better results in both image completeness distance and image coherence distance than other methods.
2012, 34(6): 1511-1515.
doi: 10.3724/SP.J.1146.2011.00704
Abstract:
Due to the capability of solving the band crowd and good performance in electromagnetic countermeasure, passive radar using illuminators of opportunity is an active research area in recent years. Taking into account the fact that most illuminators of opportunity transmit narrowband or ultra-narrowband Continuous Wave (CW) waveforms, a novel image formation method is presented for passive synthetic aperture radar (SAR) using narrowband CW waveforms. First, the data model is built up for imaging by correlating the signals received from different airborne receivers. Then an image formation method is presented using the Filtered-BackProjection (FBP) technique. Various potential factors that influence the quality of the imaging are analyzed. Finally, numerical simulations demonstrate the performance of the echo-correlation-and-FBP based passive SAR imaging method. This method does not require a priori knowledge about the transmitter locations and is applicable to the case where the radars are travelling along arbitrary flight trajectories.
Due to the capability of solving the band crowd and good performance in electromagnetic countermeasure, passive radar using illuminators of opportunity is an active research area in recent years. Taking into account the fact that most illuminators of opportunity transmit narrowband or ultra-narrowband Continuous Wave (CW) waveforms, a novel image formation method is presented for passive synthetic aperture radar (SAR) using narrowband CW waveforms. First, the data model is built up for imaging by correlating the signals received from different airborne receivers. Then an image formation method is presented using the Filtered-BackProjection (FBP) technique. Various potential factors that influence the quality of the imaging are analyzed. Finally, numerical simulations demonstrate the performance of the echo-correlation-and-FBP based passive SAR imaging method. This method does not require a priori knowledge about the transmitter locations and is applicable to the case where the radars are travelling along arbitrary flight trajectories.
2012, 34(6): 1516-1520.
doi: 10.3724/SP.J.1146.2011.01114
Abstract:
Consensus problems for mobile multi-agent systems are very practically significant in complicated kinetics systems. A consensus protocol for the first-order and the second-order discrete-time system with diverse delays in directed communication network topology is proposed. Based on the protocol, the discrete transfer function is obtained. Then generalized Nyquist criterion and Gershgorin disc method is applied to analysis the system in frequency domain. A sufficient condition is deduced for consistency convergence of the transfer function, and concrete constraint range of communication delays is given out. The simulation results show the correctness of the proposed ideas.
Consensus problems for mobile multi-agent systems are very practically significant in complicated kinetics systems. A consensus protocol for the first-order and the second-order discrete-time system with diverse delays in directed communication network topology is proposed. Based on the protocol, the discrete transfer function is obtained. Then generalized Nyquist criterion and Gershgorin disc method is applied to analysis the system in frequency domain. A sufficient condition is deduced for consistency convergence of the transfer function, and concrete constraint range of communication delays is given out. The simulation results show the correctness of the proposed ideas.
2012, 34(6): 1521-1524.
doi: 10.3724/SP.J.1146.2011.01221
Abstract:
Based on experimental analysis, this paper firstly find that the throughput of concurrent partial path transfer is better than that of concurrent all path transfer when there are more difference of path characteristic parameters among paths in Concurrent Multipath Transfer (CMT) of Internet end-to-end transport layer, so, the necessity and influence of transmission path selection in Concurrent Multipath Transfer environment are discussed. Then, fitting relations to path throughput with a variety of loss rate and delay are establish. As path throughput to path welght, a transmission path selection strategy based on path weight (i.e., CMT-PW) is proposed, the realization of algorithm and pseudocode is given. The simulation reveals that CMT-PW policy outperforms than traditional CMT policy.
Based on experimental analysis, this paper firstly find that the throughput of concurrent partial path transfer is better than that of concurrent all path transfer when there are more difference of path characteristic parameters among paths in Concurrent Multipath Transfer (CMT) of Internet end-to-end transport layer, so, the necessity and influence of transmission path selection in Concurrent Multipath Transfer environment are discussed. Then, fitting relations to path throughput with a variety of loss rate and delay are establish. As path throughput to path welght, a transmission path selection strategy based on path weight (i.e., CMT-PW) is proposed, the realization of algorithm and pseudocode is given. The simulation reveals that CMT-PW policy outperforms than traditional CMT policy.
2012, 34(6): 1483-1488.
doi: 10.3724/SP.J.1146.2011.01043
Abstract:
A new microstrip vertical transition based on the via in multilayered printed circuit board is proposed for Ultra Wide-Band (UWB) communication. The planar electromagnetic band-gap unit is etched on power plane around the vias to suppress resonance and reduce self impedance of the Power Distribution Network (PDN), and then transmission performance is improved. The interconnect structure is decoupled into microstrip mode and power ground pair mode which coupled in the vias. The network analysis method is used to estimate system performance efficiently. The simulated and measured results show that the proposed transition has an insertion loss of 0.4 dB across the frequency band 3.1-10.6 GHz. By comparison with the method of adding shorting vias between PDN, the proposed structure in this paper benefits of low cost for one less routing layer with considerable performance.
A new microstrip vertical transition based on the via in multilayered printed circuit board is proposed for Ultra Wide-Band (UWB) communication. The planar electromagnetic band-gap unit is etched on power plane around the vias to suppress resonance and reduce self impedance of the Power Distribution Network (PDN), and then transmission performance is improved. The interconnect structure is decoupled into microstrip mode and power ground pair mode which coupled in the vias. The network analysis method is used to estimate system performance efficiently. The simulated and measured results show that the proposed transition has an insertion loss of 0.4 dB across the frequency band 3.1-10.6 GHz. By comparison with the method of adding shorting vias between PDN, the proposed structure in this paper benefits of low cost for one less routing layer with considerable performance.
