Email alert
2016 Vol. 38, No. 11
Display Method:
2016, 38(11): 2701-2708.
doi: 10.11999/JEIT160011
Abstract:
In this paper, a multi-task jointly sparse spectral unmixing method based on spectral similarity measure of hyperspectral imagery is proposed, which is a refinement of collaborative sparse spectral unmixing method. First, a threshold value is obtained through the statistical characters of some random selected neighboring pixels in hypersepctral image. Second, all pixels of hyperspectral image are grouped by a spectral similarity measure and the threshold value. Then, a multi-task jointly sparse optimization problem is constructed and solved for the grouped pixels, and the abundance coefficients are obtained finally. Experimentals results on synthetic and real hyperspectral image demonstrate the effectiveness of the proposed approach.
In this paper, a multi-task jointly sparse spectral unmixing method based on spectral similarity measure of hyperspectral imagery is proposed, which is a refinement of collaborative sparse spectral unmixing method. First, a threshold value is obtained through the statistical characters of some random selected neighboring pixels in hypersepctral image. Second, all pixels of hyperspectral image are grouped by a spectral similarity measure and the threshold value. Then, a multi-task jointly sparse optimization problem is constructed and solved for the grouped pixels, and the abundance coefficients are obtained finally. Experimentals results on synthetic and real hyperspectral image demonstrate the effectiveness of the proposed approach.
2016, 38(11): 2709-2714.
doi: 10.11999/JEIT151439
Abstract:
To improve the compression ratio of lossless compression scheme based on prediction, a lossless compression scheme for hyperspectral images using K-means Clustering method and Conventional Recursive Least-Squares (C-CRLS) predictor is presented in this paper. The proposed scheme first clusters the spectral data into clusters according to their spectra using the famous K-means clustering method. Then, the proposed scheme calculates the preliminary estimates to form the input vector of the conventional recursive least-squares predictor. Finally, after prediction, the prediction residuals are sent to the arithmetic coder. Experiments on the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) 2006 hyperspectral images show that the proposed scheme yields an average compression ratio of 4.63, 2.82, and 4.77 on the 16-bit calibrated images, the 16-bit uncalibrated images, and the 12-bit uncalibrated images, respectively. Experimental results demonstrate that the proposed scheme outperforms other current state-of-the-art schemes for hyperspectral images that have been previously reported.
To improve the compression ratio of lossless compression scheme based on prediction, a lossless compression scheme for hyperspectral images using K-means Clustering method and Conventional Recursive Least-Squares (C-CRLS) predictor is presented in this paper. The proposed scheme first clusters the spectral data into clusters according to their spectra using the famous K-means clustering method. Then, the proposed scheme calculates the preliminary estimates to form the input vector of the conventional recursive least-squares predictor. Finally, after prediction, the prediction residuals are sent to the arithmetic coder. Experiments on the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) 2006 hyperspectral images show that the proposed scheme yields an average compression ratio of 4.63, 2.82, and 4.77 on the 16-bit calibrated images, the 16-bit uncalibrated images, and the 12-bit uncalibrated images, respectively. Experimental results demonstrate that the proposed scheme outperforms other current state-of-the-art schemes for hyperspectral images that have been previously reported.
2016, 38(11): 2715-2723.
doi: 10.11999/JEIT160052
Abstract:
The high dimensions of hyperspectral remote sensing images will cause the redundancy of information and complexity of data processing, which also brings tremendous computing workload and damages application accuracy. Therefore, before the analysis of hyperspectral image processing, it is necessary to reduce the high dimensions of hyperspectral data. The Sensitivity Analysis (SA) of artificial neural network can be used in dimension reduction of the model. Now the Sensitivity Analysis of artificial neural network is applied to dimension reduction for hyperspectral remote sensing images in the paper. First of all, all bands are divided into several groups as long as a lower correlation exists between adjacent bands. Furthermore, Differential Evolution (DE) algorithm is used for optimizing neural network structure. Moreover, the bands which make small contribution will be given up based on Ruck sensitivity analysis method. Finally, experiments are conducted with AVIRIS images. The results show that the proposed method can get high classification accuracy of 85.83% at small training samples, 0.31% higher than the best one among other similar methods of dimension reduction and classification.
The high dimensions of hyperspectral remote sensing images will cause the redundancy of information and complexity of data processing, which also brings tremendous computing workload and damages application accuracy. Therefore, before the analysis of hyperspectral image processing, it is necessary to reduce the high dimensions of hyperspectral data. The Sensitivity Analysis (SA) of artificial neural network can be used in dimension reduction of the model. Now the Sensitivity Analysis of artificial neural network is applied to dimension reduction for hyperspectral remote sensing images in the paper. First of all, all bands are divided into several groups as long as a lower correlation exists between adjacent bands. Furthermore, Differential Evolution (DE) algorithm is used for optimizing neural network structure. Moreover, the bands which make small contribution will be given up based on Ruck sensitivity analysis method. Finally, experiments are conducted with AVIRIS images. The results show that the proposed method can get high classification accuracy of 85.83% at small training samples, 0.31% higher than the best one among other similar methods of dimension reduction and classification.
2016, 38(11): 2724-2730.
doi: 10.11999/JEIT160104
Abstract:
At present, the amount of data collection of microwave radiometric imaging system in one snapshot is massive, so it is difficult to achieve the high spatial resolution by conventional microwave radiation imaging method based on the Nyquist sampling. According to the situations of microwave radiation interferometry conducted in the frequency domain, super sparse interferometry is adopted based on the optimal random Fourier sampling to sparsely project microwave radiation image, reducing the amount of data collection. Considering that the microwave radiation image has the character of compressibility in the total variation and microwave domain, the model of microwave radiation image reconstruction method is proposed based on the learning dictionary of mixed sparse basis of total variation and the wavelet, and the microwave radiation image is reconstructed by the Bregman and alternate direction method. The simulation results show that the proposed algorithm is better than the DLMRI algorithm and GradDLRec algorithm from two aspects of image reconstruction and noise sensitivity.
At present, the amount of data collection of microwave radiometric imaging system in one snapshot is massive, so it is difficult to achieve the high spatial resolution by conventional microwave radiation imaging method based on the Nyquist sampling. According to the situations of microwave radiation interferometry conducted in the frequency domain, super sparse interferometry is adopted based on the optimal random Fourier sampling to sparsely project microwave radiation image, reducing the amount of data collection. Considering that the microwave radiation image has the character of compressibility in the total variation and microwave domain, the model of microwave radiation image reconstruction method is proposed based on the learning dictionary of mixed sparse basis of total variation and the wavelet, and the microwave radiation image is reconstructed by the Bregman and alternate direction method. The simulation results show that the proposed algorithm is better than the DLMRI algorithm and GradDLRec algorithm from two aspects of image reconstruction and noise sensitivity.
2016, 38(11): 2731-2738.
doi: 10.11999/JEIT160195
Abstract:
The Sea State Bias (SSB) is an important source of error in satellite altimetry. Operational SSB correction models are based on the altimeter-measured wind speed (U) and Significant Wave Height (SWH). This paper presents a new method to estimate the SSB from the crossover differences using a three-dimensional nonparametric model based on U, SWH, and the Mean Wave Period (MWP). Evaluated by the separate annual data sets from 2009 to 2011, the SSB values estimated with the presented method can decrease the variance of the crossover Sea Surface Height (SSH) differences by 1.64 cm2, or 1.28 cm RMS, and decrease the variance of the Sea Level Anomalies (SLA) by 0.92 cm2, or 0.96 cm RMS in comparison to the SSB values in the Geophysical Data Records (GDR) of Jason-2. It is of great significance for improving the precision of altimeter products.
The Sea State Bias (SSB) is an important source of error in satellite altimetry. Operational SSB correction models are based on the altimeter-measured wind speed (U) and Significant Wave Height (SWH). This paper presents a new method to estimate the SSB from the crossover differences using a three-dimensional nonparametric model based on U, SWH, and the Mean Wave Period (MWP). Evaluated by the separate annual data sets from 2009 to 2011, the SSB values estimated with the presented method can decrease the variance of the crossover Sea Surface Height (SSH) differences by 1.64 cm2, or 1.28 cm RMS, and decrease the variance of the Sea Level Anomalies (SLA) by 0.92 cm2, or 0.96 cm RMS in comparison to the SSB values in the Geophysical Data Records (GDR) of Jason-2. It is of great significance for improving the precision of altimeter products.
2016, 38(11): 2739-2744.
doi: 10.11999/JEIT160051
Abstract:
In the electronic system, it occasionally needs to accomplish the transmission and reception of signal in the same platform, at this time, certain transceiver isolation and signal cancellation processing technology are usually adopted. Relative to non-simultaneous transmission and reception, simultaneous transmission and reception makes more influence on electronic warfare area. This paper takes the situation of receiving the radar signal and interfering in itself for example, makes the research of the effect of simultaneous transmission and reception on electronic warfare area. The conception of total isolation of comprehensive transceiver antenna and follow-up digital cancellation is proposed in the paper, and the factors that affect the total isolation are analyzed, then the mathematical model is set up, and the correlative formulas are deduced. In the end, the influence of total isolation on the reconnaissance function area, jamming exposure area and comprehensive electronic warfare area are analyzed by simulation.
In the electronic system, it occasionally needs to accomplish the transmission and reception of signal in the same platform, at this time, certain transceiver isolation and signal cancellation processing technology are usually adopted. Relative to non-simultaneous transmission and reception, simultaneous transmission and reception makes more influence on electronic warfare area. This paper takes the situation of receiving the radar signal and interfering in itself for example, makes the research of the effect of simultaneous transmission and reception on electronic warfare area. The conception of total isolation of comprehensive transceiver antenna and follow-up digital cancellation is proposed in the paper, and the factors that affect the total isolation are analyzed, then the mathematical model is set up, and the correlative formulas are deduced. In the end, the influence of total isolation on the reconnaissance function area, jamming exposure area and comprehensive electronic warfare area are analyzed by simulation.
2016, 38(11): 2745-2752.
doi: 10.11999/JEIT160050
Abstract:
This paper investigates the joint estimation of Time Difference Of Arrival (TDOA) and Frequency Difference Of Arrival (FDOA) in passive location system, where the true value of the reference signal is unknown. A novel Maximum Likelihood (ML) estimator of TDOA and FDOA is constructed, and Markov Chain Monte Carlo (MCMC) method is applied to finding the global maximum of likelihood function by generating the realizations of TDOA and FDOA. Unlike the Cross Ambiguity Function (CAF) algorithm or the Expectation Maximization (EM) algorithm, the proposed algorithm can also estimate the TDOA and FDOA of non-integer multiple of the sampling interval and has no dependence on the initial estimate. The Cramer Rao Lower Bound (CRLB) is also derived. Simulation results show that, the proposed algorithm outperforms the CAF and EM algorithm for different SNR conditions with higher accuracy and lower computational complexity.
This paper investigates the joint estimation of Time Difference Of Arrival (TDOA) and Frequency Difference Of Arrival (FDOA) in passive location system, where the true value of the reference signal is unknown. A novel Maximum Likelihood (ML) estimator of TDOA and FDOA is constructed, and Markov Chain Monte Carlo (MCMC) method is applied to finding the global maximum of likelihood function by generating the realizations of TDOA and FDOA. Unlike the Cross Ambiguity Function (CAF) algorithm or the Expectation Maximization (EM) algorithm, the proposed algorithm can also estimate the TDOA and FDOA of non-integer multiple of the sampling interval and has no dependence on the initial estimate. The Cramer Rao Lower Bound (CRLB) is also derived. Simulation results show that, the proposed algorithm outperforms the CAF and EM algorithm for different SNR conditions with higher accuracy and lower computational complexity.
2016, 38(11): 2753-2759.
doi: 10.11999/JEIT160125
Abstract:
Traditional design methods of two-dimensional large-scale filter banks suffer from high-complexity. This paper presents an algorithm to design two-dimensional double-prototype fully oversampled Discrete Fourier Transform (DFT) modulated filter bank with Nearly Perfect Reconstruction (NPR). The algorithm is based on bi-iterative scheme, where the design issue is formulated into an unconstrained optimization issue whose objective function is the weighted sum of the transfer distortion and the aliasing distortion of the filter bank, and the stopband energy of the Prototype Filters (PFs). By exploiting the gradient information, the optimization problem can be efficiently solved by utilizing the bi-iterative scheme. The matrix inverse identity and the fast algorithm for Toeplitz-block Toeplitz matrix inversion are employed to dramatically reduce the computational cost of the iterative procedure. The theoretical analysis and numerical experiments are carried out to show that compared with the existing methods, the new algorithm possesses much lower computational cost and can be used to design large-scale two-dimensional filter bank with better overall performance.
Traditional design methods of two-dimensional large-scale filter banks suffer from high-complexity. This paper presents an algorithm to design two-dimensional double-prototype fully oversampled Discrete Fourier Transform (DFT) modulated filter bank with Nearly Perfect Reconstruction (NPR). The algorithm is based on bi-iterative scheme, where the design issue is formulated into an unconstrained optimization issue whose objective function is the weighted sum of the transfer distortion and the aliasing distortion of the filter bank, and the stopband energy of the Prototype Filters (PFs). By exploiting the gradient information, the optimization problem can be efficiently solved by utilizing the bi-iterative scheme. The matrix inverse identity and the fast algorithm for Toeplitz-block Toeplitz matrix inversion are employed to dramatically reduce the computational cost of the iterative procedure. The theoretical analysis and numerical experiments are carried out to show that compared with the existing methods, the new algorithm possesses much lower computational cost and can be used to design large-scale two-dimensional filter bank with better overall performance.
2016, 38(11): 2760-2766.
doi: 10.11999/JEIT160042
Abstract:
For the problem of Time Division Data Modulation-Binary Offset Carrier (TDDM-BOC) modulation signal under low Signal-to-Noise Ratio (SNR), including blind estimation of the combination code sequence and information sequence, a revised method using Singular Value Decomposition (SVD) is proposed. Firstly, the received signal is divided into double-symbol-period-length temporal vectors, with one-symbol-period overlapping, accumulates of these vectors one by one to form the signal matrix. Then, an operation of SVD may be applied to the observation matrix, and the estimation of the combination code sequence is obtained based on the left singular vector. At the same time the information sequence can be estimated through the right singular vector. The simulation results show that the proposed method has accurate estimation performance for the signal at the low SNR. It has a certain reference value for engaging in satellite navigation receiver design.
For the problem of Time Division Data Modulation-Binary Offset Carrier (TDDM-BOC) modulation signal under low Signal-to-Noise Ratio (SNR), including blind estimation of the combination code sequence and information sequence, a revised method using Singular Value Decomposition (SVD) is proposed. Firstly, the received signal is divided into double-symbol-period-length temporal vectors, with one-symbol-period overlapping, accumulates of these vectors one by one to form the signal matrix. Then, an operation of SVD may be applied to the observation matrix, and the estimation of the combination code sequence is obtained based on the left singular vector. At the same time the information sequence can be estimated through the right singular vector. The simulation results show that the proposed method has accurate estimation performance for the signal at the low SNR. It has a certain reference value for engaging in satellite navigation receiver design.
2016, 38(11): 2767-2773.
doi: 10.11999/JEIT160093
Abstract:
An algorithm of Directions Of Arrival (DOA) estimation based on the state-space method is proposed to deal with the problem of estimating DOA of multiple source signals from a single observation vector of distributed array. Hankle matrixes are firstly constructed by using the single snapshot of every subarray element. Then low accuracy and unambiguous DOA estimations are obtained by the single subarray, while high accuracy and ambiguous DOA estimations are obtained by the distributed array. Finally, high accuracy and unambiguous DOA estimations are obtained by using automatic pairing decorrelating. This algorithm has no relations with the correlation signals and can fully take advantage of the large aperture to acquire high DOA estimation. Computer simulation results confirm the effectiveness of the proposed algorithm.
An algorithm of Directions Of Arrival (DOA) estimation based on the state-space method is proposed to deal with the problem of estimating DOA of multiple source signals from a single observation vector of distributed array. Hankle matrixes are firstly constructed by using the single snapshot of every subarray element. Then low accuracy and unambiguous DOA estimations are obtained by the single subarray, while high accuracy and ambiguous DOA estimations are obtained by the distributed array. Finally, high accuracy and unambiguous DOA estimations are obtained by using automatic pairing decorrelating. This algorithm has no relations with the correlation signals and can fully take advantage of the large aperture to acquire high DOA estimation. Computer simulation results confirm the effectiveness of the proposed algorithm.
2016, 38(11): 2774-2779.
doi: 10.11999/JEIT160133
Abstract:
Distance-weighted coefficients between unlabeled sample point and its nearest neighbors belonging to same sample set are determined subjectively in the Pseudo Nearest Neighbor (PNN) classification algorithm, which makes it difficult to obtain optimal distance-weighted value. In this paper, an adaptive pseudo neighbor classification algorithm based on BP neural network is proposed. Firstly, the distance-weighted values between unlabeled sample point and its neighbors lying in the same sample set are regarded as the input of BP neural network. Secondly, the corresponding distance-weighted values are adaptively determined according to the mapping between the inputs and outputs of BP neural network. Finally, the classification of unlabeled sample point is judged by the outputs of BP neural network. Experimental results show that the proposed approach adaptively adjusts the distance-weighted coefficients. Moreover, the classification accuracy can be effectively improved.
Distance-weighted coefficients between unlabeled sample point and its nearest neighbors belonging to same sample set are determined subjectively in the Pseudo Nearest Neighbor (PNN) classification algorithm, which makes it difficult to obtain optimal distance-weighted value. In this paper, an adaptive pseudo neighbor classification algorithm based on BP neural network is proposed. Firstly, the distance-weighted values between unlabeled sample point and its neighbors lying in the same sample set are regarded as the input of BP neural network. Secondly, the corresponding distance-weighted values are adaptively determined according to the mapping between the inputs and outputs of BP neural network. Finally, the classification of unlabeled sample point is judged by the outputs of BP neural network. Experimental results show that the proposed approach adaptively adjusts the distance-weighted coefficients. Moreover, the classification accuracy can be effectively improved.
2016, 38(11): 2780-2786.
doi: 10.11999/JEIT160099
Abstract:
The constant modulus algorithm is widely used in the blind equalization of wireless communication system, in order to reduce further the steady-state error, it is usually combined with the decision directed least mean square algorithm. The traditional dual mode blind equalization algorithm can achieve the hard switching of the two algorithms by setting the threshold value artificially. The rationality of the switch can not be guaranteed, it also can not fully highlight the advantages of dual mode switching. In this paper, the structure of the convex combination is used to achieve the switching between two modes and can adaptively choose the switching time. It improves the convergence rate and reduces the steady-state error by modifying the algorithm and normalizing the mixed parameter. In addition, the steady-state performance is derived and analyzed. Simulation results demonstrate that the performance of the model is consistent with the results, the effect of parameter normalization is obvious. Compared with other similar dual mode switching algorithms, it has better performance.
The constant modulus algorithm is widely used in the blind equalization of wireless communication system, in order to reduce further the steady-state error, it is usually combined with the decision directed least mean square algorithm. The traditional dual mode blind equalization algorithm can achieve the hard switching of the two algorithms by setting the threshold value artificially. The rationality of the switch can not be guaranteed, it also can not fully highlight the advantages of dual mode switching. In this paper, the structure of the convex combination is used to achieve the switching between two modes and can adaptively choose the switching time. It improves the convergence rate and reduces the steady-state error by modifying the algorithm and normalizing the mixed parameter. In addition, the steady-state performance is derived and analyzed. Simulation results demonstrate that the performance of the model is consistent with the results, the effect of parameter normalization is obvious. Compared with other similar dual mode switching algorithms, it has better performance.
2016, 38(11): 2787-2794.
doi: 10.11999/JEIT160161
Abstract:
A novel image forensic algorithm against contrast modification based on superpixel and histogram of run length is proposed. In the proposed algorithm, images are firstly divided by superpixel, then run length histogram features of each block are extracted, and those of different orientation are subsequently merged. After normalization of the prior features, the leaps in the histogram are calculated numerically. Lastly, the generated features of blocks are trained by Support Vector Machin (SVM) classifier. Large amounts of experiments show that, the proposed algorithm has low cost of computation complexity. And compared with some state-of-the-art algorithms, it has better performance with many test databases. Furthermore, the proposed algorithm can not only located the tempered area, but also can exactly describe the shape of tempered area.
A novel image forensic algorithm against contrast modification based on superpixel and histogram of run length is proposed. In the proposed algorithm, images are firstly divided by superpixel, then run length histogram features of each block are extracted, and those of different orientation are subsequently merged. After normalization of the prior features, the leaps in the histogram are calculated numerically. Lastly, the generated features of blocks are trained by Support Vector Machin (SVM) classifier. Large amounts of experiments show that, the proposed algorithm has low cost of computation complexity. And compared with some state-of-the-art algorithms, it has better performance with many test databases. Furthermore, the proposed algorithm can not only located the tempered area, but also can exactly describe the shape of tempered area.
2016, 38(11): 2795-2802.
doi: 10.11999/JEIT160013
Abstract:
Considering that the existing high-order models have limitations in forecast range and accuracy, a heuristic adaptive-order intuitionistic fuzzy time series forecasting model is built with the combination of the intuitionistic fuzzy sets theory. In this model, a direct fuzzy clustering algorithm is used to partition the universe of discourse into unequal intervals. The traditional method of ascertaining the membership and non-membership functions of intuitionistic fuzzy set are also modified to fit the intuitionistic fuzzy time series data. On these basis, variable high-order forecasting rules are established and the prior knowledge of tendency is used in defuzzification to extend the forecasting range. At last, contrast experiments on the enrollments of the University of Alabama and the daily average temperature of Beijing are carried out. The results show that the new model has a clear advantage of improving the forecast accuracy.
Considering that the existing high-order models have limitations in forecast range and accuracy, a heuristic adaptive-order intuitionistic fuzzy time series forecasting model is built with the combination of the intuitionistic fuzzy sets theory. In this model, a direct fuzzy clustering algorithm is used to partition the universe of discourse into unequal intervals. The traditional method of ascertaining the membership and non-membership functions of intuitionistic fuzzy set are also modified to fit the intuitionistic fuzzy time series data. On these basis, variable high-order forecasting rules are established and the prior knowledge of tendency is used in defuzzification to extend the forecasting range. At last, contrast experiments on the enrollments of the University of Alabama and the daily average temperature of Beijing are carried out. The results show that the new model has a clear advantage of improving the forecast accuracy.
2016, 38(11): 2803-2810.
doi: 10.11999/JEIT160122
Abstract:
Under the framework of the Bayesian inference, tracking methods based on PCA subspace and L2-norm minimization can deal with some complex appearance changes in the video scene successfully. However, they are prone to drifting or failure when the target object undergoes pose variation or rotation. To deal with this problem, a fast visual tracking method is proposed based on L2-norm minimization and compressed Haar-like features matching. The proposed method not only removes square templates, but also presents a simple but effective observation likelihood, and its robustness to pose variation and rotation is strengthened by Haar-like features matching. Compared with other popular method, the proposed method has stronger robustness to abnormal changes (e.g. heavy occlusion, drastic illumination change, abrupt motion, pose variation and rotation, etc). Furthermore, it runs fast with a speed of about 29 frames/s.
Under the framework of the Bayesian inference, tracking methods based on PCA subspace and L2-norm minimization can deal with some complex appearance changes in the video scene successfully. However, they are prone to drifting or failure when the target object undergoes pose variation or rotation. To deal with this problem, a fast visual tracking method is proposed based on L2-norm minimization and compressed Haar-like features matching. The proposed method not only removes square templates, but also presents a simple but effective observation likelihood, and its robustness to pose variation and rotation is strengthened by Haar-like features matching. Compared with other popular method, the proposed method has stronger robustness to abnormal changes (e.g. heavy occlusion, drastic illumination change, abrupt motion, pose variation and rotation, etc). Furthermore, it runs fast with a speed of about 29 frames/s.
2016, 38(11): 2811-2818.
doi: 10.11999/JEIT160009
Abstract:
The nuclear norm is used to replace the rank function in the subspace clustering algorithm based on low rank representation, it can not estimate the rank of the matrix effectively and it is sensitive to Gauss noise. In this paper, a novel algorithm is proposed to improve the accuracy and maintain its stability under the Gauss noise. When building the objective function, the nuclear norm and Forbenius norm of coefficient matrix are used as the regularization terms, after a strong convex regularizer over singular values of coefficient matrix, an inexact augmented Lagrange multiplier method is utilized to solve the problem. Finally, the affinity matrix is acquired by post-processing of coefficient matrix and the classical spectral clustering method is employed to clustering. The experimental comparison results between the state-of-the-art algorithms on synthetic data, Extended Yale B and PIE datasets demonstrate the effectiveness of the proposed improved method and the robustness to Gauss noise.
The nuclear norm is used to replace the rank function in the subspace clustering algorithm based on low rank representation, it can not estimate the rank of the matrix effectively and it is sensitive to Gauss noise. In this paper, a novel algorithm is proposed to improve the accuracy and maintain its stability under the Gauss noise. When building the objective function, the nuclear norm and Forbenius norm of coefficient matrix are used as the regularization terms, after a strong convex regularizer over singular values of coefficient matrix, an inexact augmented Lagrange multiplier method is utilized to solve the problem. Finally, the affinity matrix is acquired by post-processing of coefficient matrix and the classical spectral clustering method is employed to clustering. The experimental comparison results between the state-of-the-art algorithms on synthetic data, Extended Yale B and PIE datasets demonstrate the effectiveness of the proposed improved method and the robustness to Gauss noise.
2016, 38(11): 2819-2825.
doi: 10.11999/JEIT160114
Abstract:
As a typical implementation of the Probability Hypothesis Density (PHD) filter, Particle PHD (P-PHD) is suitable for highly nonlinear systems and widely used in Multi-Target Tracking (MTT). However, the resampling in P-PHD filter, recommended to avoid particle degeneracy, introduces the problem of diversity loss among the particles, namely particle impoverishment problem. To solve the problem and improve the performance of the P-PHD filter, a novel filter based on stochastic perturbation re-sampling is proposed. First, a comprehensive analysis on the particle impoverishment problem of P-PHD filter is presented. Then for the purpose of keeping the particle diversity, a new stochastic perturbation re-sampling algorithm is developed, which generates new particles according to the position and duplicating times of the original particles, and removes some excessive copied particles. Finally, the re-sampling algorithm is integrated into the P-PHD filter framework and a Stochastic Perturbation Particle PHD (SPP-PHD) filter is proposed. Numerical examples show that the proposed filter can overcome the particle impoverishment problem and improve the estimation performance on the premise of not significantly improving the simulation time.
As a typical implementation of the Probability Hypothesis Density (PHD) filter, Particle PHD (P-PHD) is suitable for highly nonlinear systems and widely used in Multi-Target Tracking (MTT). However, the resampling in P-PHD filter, recommended to avoid particle degeneracy, introduces the problem of diversity loss among the particles, namely particle impoverishment problem. To solve the problem and improve the performance of the P-PHD filter, a novel filter based on stochastic perturbation re-sampling is proposed. First, a comprehensive analysis on the particle impoverishment problem of P-PHD filter is presented. Then for the purpose of keeping the particle diversity, a new stochastic perturbation re-sampling algorithm is developed, which generates new particles according to the position and duplicating times of the original particles, and removes some excessive copied particles. Finally, the re-sampling algorithm is integrated into the P-PHD filter framework and a Stochastic Perturbation Particle PHD (SPP-PHD) filter is proposed. Numerical examples show that the proposed filter can overcome the particle impoverishment problem and improve the estimation performance on the premise of not significantly improving the simulation time.
2016, 38(11): 2826-2831.
doi: 10.11999/JEIT160150
Abstract:
One significant advantage of rate selection algorithms based on bandwidth estimation is the high bandwidth utilization rate. They are, however, vulnerable to network bandwidth fluctuations, leading to appearance of rate instantaneous peak value and hence wasting unnecessary bandwidth consumption. To tackle the problem above, this paper proposes a novel rate selection algorithm based on the contrast sensitivity of human eyes, where in the client eyes cutoff spatial frequency under the current viewing conditions is calculated by using the human contrast sensitivity model. The algorithm selects the rate of video fragment which has the minimum absolute difference value to the spatial frequency computed, stored in server as the target rate. Compared with those methods for calculating the target rate based on bandwidth estimation and testing target rate in different angles, the proposed method gets the ladder diagrams of rate calculation of both methods. Experimental results demonstrate that the proposed algorithm is able to save a considerable amount of bandwidth without the loss of video quality, with viewing angle from 5 to 15.
One significant advantage of rate selection algorithms based on bandwidth estimation is the high bandwidth utilization rate. They are, however, vulnerable to network bandwidth fluctuations, leading to appearance of rate instantaneous peak value and hence wasting unnecessary bandwidth consumption. To tackle the problem above, this paper proposes a novel rate selection algorithm based on the contrast sensitivity of human eyes, where in the client eyes cutoff spatial frequency under the current viewing conditions is calculated by using the human contrast sensitivity model. The algorithm selects the rate of video fragment which has the minimum absolute difference value to the spatial frequency computed, stored in server as the target rate. Compared with those methods for calculating the target rate based on bandwidth estimation and testing target rate in different angles, the proposed method gets the ladder diagrams of rate calculation of both methods. Experimental results demonstrate that the proposed algorithm is able to save a considerable amount of bandwidth without the loss of video quality, with viewing angle from 5 to 15.
2016, 38(11): 2832-2839.
doi: 10.11999/JEIT160169
Abstract:
Some existing color image adaptive watermarking algorithms do not fully utilize the color information in the adaptive process, or do not consider the holistic property of the components of a color host image. To overcome these drawbacks, this paper proposes a color image adaptive watermarking algorithm based on Fractional Quaternion Fourier Transform (FrQFT). Firstly, the texture, edge and color tone features of the blocks of the host image are extracted using the Human Vision System (HVS). After that, the embedding strength values of the blocks suitable to watermark embedding are set adaptively according to the extracted feature. Finally, the quantization index modulation and the multiple redundant embedding strategy are used to insert the watermark in the FrQFT domain with the adaptive strength. Experimental results show that the proposed algorithm is superior over the existing algorithm using Quaternion Fourier Transform (QFT) and the algorithm based on Fractional Fourier Transform (FrFT).
Some existing color image adaptive watermarking algorithms do not fully utilize the color information in the adaptive process, or do not consider the holistic property of the components of a color host image. To overcome these drawbacks, this paper proposes a color image adaptive watermarking algorithm based on Fractional Quaternion Fourier Transform (FrQFT). Firstly, the texture, edge and color tone features of the blocks of the host image are extracted using the Human Vision System (HVS). After that, the embedding strength values of the blocks suitable to watermark embedding are set adaptively according to the extracted feature. Finally, the quantization index modulation and the multiple redundant embedding strategy are used to insert the watermark in the FrQFT domain with the adaptive strength. Experimental results show that the proposed algorithm is superior over the existing algorithm using Quaternion Fourier Transform (QFT) and the algorithm based on Fractional Fourier Transform (FrFT).
2016, 38(11): 2840-2847.
doi: 10.11999/JEIT160209
Abstract:
ElectroEncephaloGram (EEG) data are often contaminated by various electrophysiological artifacts. Among all these artifacts, removing the ones related to muscle activity is particularly challenging. In past studies, Independent Component Analysis (ICA) and Canonical Correlation Analysis (CCA), as Blind Source Separation (BSS) methods, are widely used. In this work, a new method for muscle artifact removal in EEG data using Independent Vector Analysis (IVA) is proposed. IVA utilizes both the higher-order and second-order statistics, so that it makes full use of non-Gaussianity and weak autocorrelation of the muscle artifact and has the advantages of both ICA and CCA. The proposed method is examined on a number of simulated data sets and is shown to have better performance than ICA and CCA. The proposed IVA method is able to largely suppress muscle activity and meanwhile well preserve the underlying EEG activity.
ElectroEncephaloGram (EEG) data are often contaminated by various electrophysiological artifacts. Among all these artifacts, removing the ones related to muscle activity is particularly challenging. In past studies, Independent Component Analysis (ICA) and Canonical Correlation Analysis (CCA), as Blind Source Separation (BSS) methods, are widely used. In this work, a new method for muscle artifact removal in EEG data using Independent Vector Analysis (IVA) is proposed. IVA utilizes both the higher-order and second-order statistics, so that it makes full use of non-Gaussianity and weak autocorrelation of the muscle artifact and has the advantages of both ICA and CCA. The proposed method is examined on a number of simulated data sets and is shown to have better performance than ICA and CCA. The proposed IVA method is able to largely suppress muscle activity and meanwhile well preserve the underlying EEG activity.
2016, 38(11): 2848-2855.
doi: 10.11999/JEIT160082
Abstract:
ElectroEncephaloGram (EEG) signal detection and recognition is an important diagnostic method for the epilepsy. Radial Basis Function (RBF) neural network has excellent performance on approximation and generalization, and can directly recognize EEG signals in different states. However, its transparency and interpretability are low, and it also ignore the different separabilities between different classes of data. In this paper, a classification tree based on RBF neural networks and minimax probability decision technique is proposed, using one-against-one and exclusive method and paying much attention to the different separabilities among classes. Experiments on EEG signals show that the proposed method has clear structure, strong classification ability and better interpretability.
ElectroEncephaloGram (EEG) signal detection and recognition is an important diagnostic method for the epilepsy. Radial Basis Function (RBF) neural network has excellent performance on approximation and generalization, and can directly recognize EEG signals in different states. However, its transparency and interpretability are low, and it also ignore the different separabilities between different classes of data. In this paper, a classification tree based on RBF neural networks and minimax probability decision technique is proposed, using one-against-one and exclusive method and paying much attention to the different separabilities among classes. Experiments on EEG signals show that the proposed method has clear structure, strong classification ability and better interpretability.
2016, 38(11): 2856-2863.
doi: 10.11999/JEIT160181
Abstract:
Texture factor is one of the most important characteristics in the image description. In order to describe the texture feature accurately, and enhance image distinguish ability, a method of texture image retrieval is proposed based on Dual-Tree Complex Wavelet Transform (DT-CWT) in this paper. Firstly, each sub-band coefficient is obtained by DT-CWT, because the coefficient distribution exists slight incomplete symmetrical feature, which is modeled as dual-generalized Gaussian model. Secondly, there is incomplete independent and uncertain conflict between the sub-band coefficients, therefore the Fuzzy Set and Dempster-Shafer (FS-DS) evidence theory are applied to blending the characteristics of each subband coefficients. The performance of the propose algorithm is tested on the Brodatz and color texture image library, and also compared with a variety of statistical modeling methods. The experimental results demonstrate that the proposed method can improve the average retrieval rate of the texture images effectively.
Texture factor is one of the most important characteristics in the image description. In order to describe the texture feature accurately, and enhance image distinguish ability, a method of texture image retrieval is proposed based on Dual-Tree Complex Wavelet Transform (DT-CWT) in this paper. Firstly, each sub-band coefficient is obtained by DT-CWT, because the coefficient distribution exists slight incomplete symmetrical feature, which is modeled as dual-generalized Gaussian model. Secondly, there is incomplete independent and uncertain conflict between the sub-band coefficients, therefore the Fuzzy Set and Dempster-Shafer (FS-DS) evidence theory are applied to blending the characteristics of each subband coefficients. The performance of the propose algorithm is tested on the Brodatz and color texture image library, and also compared with a variety of statistical modeling methods. The experimental results demonstrate that the proposed method can improve the average retrieval rate of the texture images effectively.
2016, 38(11): 2864-2870.
doi: 10.11999/JEIT160080
Abstract:
To solve the speed bottleneck of deformable part models in the detection process, this paper proposes a cascade deformable part model with rapid computation of feature pyramids for the detection process of the model. Because the speed of the detection is mainly determined by the two processes of the feature computation and the object location, a two-stage speedup algorithm is proposed. Firstly, sparsely-sampled feature pyramids on the scale are utilized to approximate finely-sampled multi-scale image features to speed up the process of feature computation. Then combined with the cascade algorithm in the location process, a sequence model is utilized to evaluate individual parts sequentially so as to rapidly prune most object hypotheses of small possibilities in order to speed up the process of object location. The experimental results on PASCAL VOC 2007 dataset and INRIA dataset show that the algorithm in the paper apparently speeds up the speed of detection with minor loss in detection precision.
To solve the speed bottleneck of deformable part models in the detection process, this paper proposes a cascade deformable part model with rapid computation of feature pyramids for the detection process of the model. Because the speed of the detection is mainly determined by the two processes of the feature computation and the object location, a two-stage speedup algorithm is proposed. Firstly, sparsely-sampled feature pyramids on the scale are utilized to approximate finely-sampled multi-scale image features to speed up the process of feature computation. Then combined with the cascade algorithm in the location process, a sequence model is utilized to evaluate individual parts sequentially so as to rapidly prune most object hypotheses of small possibilities in order to speed up the process of object location. The experimental results on PASCAL VOC 2007 dataset and INRIA dataset show that the algorithm in the paper apparently speeds up the speed of detection with minor loss in detection precision.
2016, 38(11): 2871-2878.
doi: 10.11999/JEIT160165
Abstract:
Many systemic diseases can cause changes of the diameters of retinal vessels and Arteriolar-to-Venular diameter Ratios (AVR), so it is of great importance to make quantitative analysis of the diameter of retinal vessels accurately in the diagnosis of the disease. An automatic method measuring the diameters of the artery and the vein and the AVR is proposed. Firstly, based on the segmentation of vascular network, the diameters of retinal vessels are measured according to the advantage of Hesse matrix for detecting line-like structure, and accurate localization of vascular direction with multi-scale analysis. Secondly, a General Regression Neural Network (GRNN) classifier is used to classify the artery and the vein points. Finally, the AVR in the Region Of Interest (ROI) is calculated. The validity of the proposed method is demonstrated by testing on the DRIVE and the REVIEW database.
Many systemic diseases can cause changes of the diameters of retinal vessels and Arteriolar-to-Venular diameter Ratios (AVR), so it is of great importance to make quantitative analysis of the diameter of retinal vessels accurately in the diagnosis of the disease. An automatic method measuring the diameters of the artery and the vein and the AVR is proposed. Firstly, based on the segmentation of vascular network, the diameters of retinal vessels are measured according to the advantage of Hesse matrix for detecting line-like structure, and accurate localization of vascular direction with multi-scale analysis. Secondly, a General Regression Neural Network (GRNN) classifier is used to classify the artery and the vein points. Finally, the AVR in the Region Of Interest (ROI) is calculated. The validity of the proposed method is demonstrated by testing on the DRIVE and the REVIEW database.
2016, 38(11): 2879-2886.
doi: 10.11999/JEIT160157
Abstract:
In order to improve the correct rate of Voice Activity Detection (VAD) in low Signal Noise Ratio (SNR) environment, the paper presents an adaptive reserved likelihood ratio VAD method, which is based on sub-band double features. The method employs sub-band auto correlate function and sub-band zero crossing rate in the process of setting reserved weight. Reserved threshold is estimated adaptively according to the passed VAD results and their sub-band feature parameters. The experiment shows its promising performance in comparison with similar algorithms, the VAD correct rate is improved by 1.2%, 7.2%, and 8.1% respectively in 10 dB, 0 dB, and -10 dB stationary white noisy environment, 1.6% and 3.4% respectively in 10 dB and 0 dB non-stationary Babble noisy environment. The method is also applied to 2.4 kbps low bit rate vocoder and the Perceptual Evaluation of Speech Quality (PESQ) is improved by 0.098~0.153 in white noisy environment, 0.157~0.186 in Babble noisy environment.
In order to improve the correct rate of Voice Activity Detection (VAD) in low Signal Noise Ratio (SNR) environment, the paper presents an adaptive reserved likelihood ratio VAD method, which is based on sub-band double features. The method employs sub-band auto correlate function and sub-band zero crossing rate in the process of setting reserved weight. Reserved threshold is estimated adaptively according to the passed VAD results and their sub-band feature parameters. The experiment shows its promising performance in comparison with similar algorithms, the VAD correct rate is improved by 1.2%, 7.2%, and 8.1% respectively in 10 dB, 0 dB, and -10 dB stationary white noisy environment, 1.6% and 3.4% respectively in 10 dB and 0 dB non-stationary Babble noisy environment. The method is also applied to 2.4 kbps low bit rate vocoder and the Perceptual Evaluation of Speech Quality (PESQ) is improved by 0.098~0.153 in white noisy environment, 0.157~0.186 in Babble noisy environment.
2016, 38(11): 2887-2892.
doi: 10.11999/JEIT160054
Abstract:
A physical layer security scheme is studied, which employs the advantage of artificial noise to improve the performance of legitimate user for multiple antenna systems using beamforming technology and artificial noise. Based on the transmitted symbols and channel coefficients, the sender determines whether or not the artificial noises are beneficial to the signal detection at the legitimate receiver. Then, beamforming vectors are designed accordingly. By taking advantage of useful noise, the signal to noise ratio at the legitimate receiver is improved effectively while that at the illegal receiver will remain the same. The bit error rate and the secrecy capacity are analyzed and simulated. The results demonstrate that the proposed scheme can improve the performance of the legal receiver and enhance secrecy capacity.
A physical layer security scheme is studied, which employs the advantage of artificial noise to improve the performance of legitimate user for multiple antenna systems using beamforming technology and artificial noise. Based on the transmitted symbols and channel coefficients, the sender determines whether or not the artificial noises are beneficial to the signal detection at the legitimate receiver. Then, beamforming vectors are designed accordingly. By taking advantage of useful noise, the signal to noise ratio at the legitimate receiver is improved effectively while that at the illegal receiver will remain the same. The bit error rate and the secrecy capacity are analyzed and simulated. The results demonstrate that the proposed scheme can improve the performance of the legal receiver and enhance secrecy capacity.
2016, 38(11): 2893-2899.
doi: 10.11999/JEIT151444
Abstract:
In order to solve the problem of blind detection of MIMO system, this paper takes maximum-likelihood sequence detection as the criterion and derives the formulas to get a model based on SemidDefinite Relaxation. The rank of SDR solution equals to the number of the transmit antennas. For the rank of SDR solution is greater than 1, a new method is proposed to approximate the solution of the original problem, which combines the eigenvector approximation method and randomization method. By setting the upper limit of objective function, the proposed method could judge the number of detection sequence adaptively and reduce constrains number and the number of solving SDR. The analysis shows that the computation complexity of proposed method has linear relationship with the number of transmit antennas. At last, simulation results indicate that compared with Rank-1 algorithm, the proposed detector could provide the same bit error performance with decrease of computation cost, and validate the linear relationship between the computation complexity and the number of transmit antennas.
In order to solve the problem of blind detection of MIMO system, this paper takes maximum-likelihood sequence detection as the criterion and derives the formulas to get a model based on SemidDefinite Relaxation. The rank of SDR solution equals to the number of the transmit antennas. For the rank of SDR solution is greater than 1, a new method is proposed to approximate the solution of the original problem, which combines the eigenvector approximation method and randomization method. By setting the upper limit of objective function, the proposed method could judge the number of detection sequence adaptively and reduce constrains number and the number of solving SDR. The analysis shows that the computation complexity of proposed method has linear relationship with the number of transmit antennas. At last, simulation results indicate that compared with Rank-1 algorithm, the proposed detector could provide the same bit error performance with decrease of computation cost, and validate the linear relationship between the computation complexity and the number of transmit antennas.
2016, 38(11): 2900-2907.
doi: 10.11999/JEIT160007
Abstract:
To solve the problem of key leakages in existing physical layer challenge-response authentication schemes, a physical layer authentication scheme based on hash method is proposed. The channel characteristics are extracted and linked with the key which can be regarded as a curve. Then a fault-tolerant hash function is employed to map the curve into a response with lower dimension. The authenticator lastly sets the threshold according to the authentication requirement and further to verify the identity of the requester. The hash function is an underdetermined system and attackers can not recover the curve according to the response. Simulation results prove the effectiveness of the scheme whose attack rate is less than10-5 while attack rates for existing schemes are almost 0.5 under the SNR of 4 dB.
To solve the problem of key leakages in existing physical layer challenge-response authentication schemes, a physical layer authentication scheme based on hash method is proposed. The channel characteristics are extracted and linked with the key which can be regarded as a curve. Then a fault-tolerant hash function is employed to map the curve into a response with lower dimension. The authenticator lastly sets the threshold according to the authentication requirement and further to verify the identity of the requester. The hash function is an underdetermined system and attackers can not recover the curve according to the response. Simulation results prove the effectiveness of the scheme whose attack rate is less than10-5 while attack rates for existing schemes are almost 0.5 under the SNR of 4 dB.
2016, 38(11): 2908-2915.
doi: 10.11999/JEIT151477
Abstract:
Shuffled-BP (SBP) decoding algorithm is a variable-node-based serial decoding algorithm, which converges faster than the original Belief-Propagation (BP) decoding algorithm. However, due to the semi-parallel processing, there is a decrease in terms of convergence speed and error performance. An Overlapped Shuffled-BP(OSBP) decoding algorithm is proposed to enhance further the performance of the Shuffled-BP algorithm. In this algorithm, more than one sub-decoders are used to execute simultaneously, every sub-decoder has different updating order from each other. Regarding each variable node, the most reliable messages are kept and used for the next iteration, thus a faster convergence can be provided. Both theoretical analysis and simulation results show that, compared with SBP algorithm, OSBP algorithm possesses a better error performance as well as a higher convergence speed and introduces no extra storage requirement. Moreover, the proposed algorithm is effective for both regular and irregular LDPC codes.
Shuffled-BP (SBP) decoding algorithm is a variable-node-based serial decoding algorithm, which converges faster than the original Belief-Propagation (BP) decoding algorithm. However, due to the semi-parallel processing, there is a decrease in terms of convergence speed and error performance. An Overlapped Shuffled-BP(OSBP) decoding algorithm is proposed to enhance further the performance of the Shuffled-BP algorithm. In this algorithm, more than one sub-decoders are used to execute simultaneously, every sub-decoder has different updating order from each other. Regarding each variable node, the most reliable messages are kept and used for the next iteration, thus a faster convergence can be provided. Both theoretical analysis and simulation results show that, compared with SBP algorithm, OSBP algorithm possesses a better error performance as well as a higher convergence speed and introduces no extra storage requirement. Moreover, the proposed algorithm is effective for both regular and irregular LDPC codes.
2016, 38(11): 2916-2921.
doi: 10.11999/JEIT160074
Abstract:
Because the matroid search algorithm is very complicated and the local matroid search algorithm can not search all optimal codes, this paper proposes a variable matroid search algorithm to search the quasi-cyclic codes by researching matroid search algorithm. The algorithm reduces the computational complexity by reducing the repeated search. Based on this algorithm, the systematic binary quasi-cyclic codes of which the rate is 1/p are constructed. With the change of integer p, the optimal codes of rate 1/p can be obtained by the generator matrix reducing or adding a loop matrix. Through experiments, two new codes of which the minimum distance is larger than the existing optimal codes are worked out, which indicate the feasibility and superiority of the algorithm.
Because the matroid search algorithm is very complicated and the local matroid search algorithm can not search all optimal codes, this paper proposes a variable matroid search algorithm to search the quasi-cyclic codes by researching matroid search algorithm. The algorithm reduces the computational complexity by reducing the repeated search. Based on this algorithm, the systematic binary quasi-cyclic codes of which the rate is 1/p are constructed. With the change of integer p, the optimal codes of rate 1/p can be obtained by the generator matrix reducing or adding a loop matrix. Through experiments, two new codes of which the minimum distance is larger than the existing optimal codes are worked out, which indicate the feasibility and superiority of the algorithm.
2016, 38(11): 2922-2930.
doi: 10.11999/JEIT160045
Abstract:
In order to deal with the explosive growth of mobile data traffic, a novel design of network architecture will be adopted in 5G. Software Defined Network (SDN) and Network Function Virtualization (NFV) are the key technologies for network transformation, which will drive the innovation of mobile communication network architecture. The deployment of Virtualized Network Function (VNF) in service chain is a critical issue in network virtualization. To overcome the ignorance of VNF sequence constraints in service chain and the characteristics of mobile business in existing literatures, an adaptive deployment method of VNF based on Viterbi algorithm is proposed. With real-time perception of the resources change of underlying nodes, the topology structure will be adjusted dynamically. Hidden Markov model is used to describe the topology information of available nodes with resources constraints in underlying network, and the service path with shortest delay is selected based on Viterbi algorithm in candidate service node. Experimental results show that the process time of service chain can be lower compared with existing algorithm. In addition, the acceptance rates of service chain requests and cost efficiency of underlying resources are also raised.
In order to deal with the explosive growth of mobile data traffic, a novel design of network architecture will be adopted in 5G. Software Defined Network (SDN) and Network Function Virtualization (NFV) are the key technologies for network transformation, which will drive the innovation of mobile communication network architecture. The deployment of Virtualized Network Function (VNF) in service chain is a critical issue in network virtualization. To overcome the ignorance of VNF sequence constraints in service chain and the characteristics of mobile business in existing literatures, an adaptive deployment method of VNF based on Viterbi algorithm is proposed. With real-time perception of the resources change of underlying nodes, the topology structure will be adjusted dynamically. Hidden Markov model is used to describe the topology information of available nodes with resources constraints in underlying network, and the service path with shortest delay is selected based on Viterbi algorithm in candidate service node. Experimental results show that the process time of service chain can be lower compared with existing algorithm. In addition, the acceptance rates of service chain requests and cost efficiency of underlying resources are also raised.
2016, 38(11): 2931-2939.
doi: 10.11999/JEIT160064
Abstract:
Attribute-Based Encryption (ABE) is often used in cloud storage to achieve fine-grained access control. In order to further protect the sensitive information of access control policy and solve the key escrow caused by the authority center generating the private key for users alone. In this paper, the attributes of access control policy are remapped to achieve its privacy. Additionally, a two-party computing protocol in which the user generates partial private key component is devised to solve the problem of key escrow. At last, the security of this scheme is proved in the standard model, and the performance analysis and experiment validation are conducted, which show that although some additional computation overhead is added for achieving the privacy of access control policy and solving the problem of key escrow, the receiver in proposed scheme has smaller computation overhead compared with the existing related schemes because most of the decryption is delegated to the storage center to carry out.
Attribute-Based Encryption (ABE) is often used in cloud storage to achieve fine-grained access control. In order to further protect the sensitive information of access control policy and solve the key escrow caused by the authority center generating the private key for users alone. In this paper, the attributes of access control policy are remapped to achieve its privacy. Additionally, a two-party computing protocol in which the user generates partial private key component is devised to solve the problem of key escrow. At last, the security of this scheme is proved in the standard model, and the performance analysis and experiment validation are conducted, which show that although some additional computation overhead is added for achieving the privacy of access control policy and solving the problem of key escrow, the receiver in proposed scheme has smaller computation overhead compared with the existing related schemes because most of the decryption is delegated to the storage center to carry out.
2016, 38(11): 2940-2947.
doi: 10.11999/JEIT160105
Abstract:
In order to provide differentiated services in information centric networking, a Differentiated Service Model based on Meta Module (DSM3) is proposed. DSM3 defines the basic network control unit as meta module, and matches different meta module combination cases to carry different business with various demand characteristics. The meta module combination process is deduced as the secondary mapping problem of business policycase combination stringsbusiness path. Then, the meta module cases of route calculation for real-time service, non-real time streaming media and user generated content are designed. Simulation results show that through a small amount of additional control overhead, DSM3 reduces the average response delay of the three kinds of business above, improves the network average cache hit rate, and supports differentiated services.
In order to provide differentiated services in information centric networking, a Differentiated Service Model based on Meta Module (DSM3) is proposed. DSM3 defines the basic network control unit as meta module, and matches different meta module combination cases to carry different business with various demand characteristics. The meta module combination process is deduced as the secondary mapping problem of business policycase combination stringsbusiness path. Then, the meta module cases of route calculation for real-time service, non-real time streaming media and user generated content are designed. Simulation results show that through a small amount of additional control overhead, DSM3 reduces the average response delay of the three kinds of business above, improves the network average cache hit rate, and supports differentiated services.
2016, 38(11): 2948-2953.
doi: 10.11999/JEIT160056
Abstract:
In the past studies, it is generally assumed that both sides of communication are in the same environment of public key cryptography, but with the development of technology and the popularity of the network, different regions may have different public key cryptographies. In order to resolve the communication security problem between heterogeneous systems, two signcryption schemes are proposed, which are used to achieve the communication security between the Public Key Infrastructure (PKI) and CertificatLess public key Cryptography (CLC) under heterogeneous systems. It is proved that the schemes have INDistinguishability against Adaptive Chosen Ciphertext Attacks (IND-CCA2) under Bilinear Diffie-Hellman Problem (BDHP) and Existential UnForgeability against adaptive Chosen Messages Attacks (EUF-CMA) under the Computational Diffie-Hellman Problem (CDHP) in the random oracle model.
In the past studies, it is generally assumed that both sides of communication are in the same environment of public key cryptography, but with the development of technology and the popularity of the network, different regions may have different public key cryptographies. In order to resolve the communication security problem between heterogeneous systems, two signcryption schemes are proposed, which are used to achieve the communication security between the Public Key Infrastructure (PKI) and CertificatLess public key Cryptography (CLC) under heterogeneous systems. It is proved that the schemes have INDistinguishability against Adaptive Chosen Ciphertext Attacks (IND-CCA2) under Bilinear Diffie-Hellman Problem (BDHP) and Existential UnForgeability against adaptive Chosen Messages Attacks (EUF-CMA) under the Computational Diffie-Hellman Problem (CDHP) in the random oracle model.
2016, 38(11): 2954-2959.
doi: 10.11999/JEIT160049
Abstract:
In the design of ultra wideband TEM horn antenna, it is difficult to combine both the character optimization in low frequency and the miniaturization of an antenna, which always limits the usage of TEM horns in some applications. For the problem of low-frequency reflection, a structure, which is referred to the principle of band-pass filters in parallel and different from the conventional ones, is proposed in this paper for separating a fraction of low-frequency electric field component on the plates linking excitation port to radiation plates. It can greatly reduce the reflection of low-frequency component at the edge of radiation plates to improve the low frequency character of the antenna. In this paper, an ultra-wideband TEM horn antenna improved with the proposed structure is presented. Finally, the comparison of the improved antenna and the original one is exhibited, which shows that the impedance bandwidth is 12.5% wider with the low-cutoff frequency decreasing to 0.1 GHz and simultaneously the port-feed efficiency is increasing by 10%. The results confirm the validity of the proposed design to optimize the low frequency character. The feasibility is also validated in the end by analyzing the impedance and the magnitude distribution of currents in different frequencies, respectively.
In the design of ultra wideband TEM horn antenna, it is difficult to combine both the character optimization in low frequency and the miniaturization of an antenna, which always limits the usage of TEM horns in some applications. For the problem of low-frequency reflection, a structure, which is referred to the principle of band-pass filters in parallel and different from the conventional ones, is proposed in this paper for separating a fraction of low-frequency electric field component on the plates linking excitation port to radiation plates. It can greatly reduce the reflection of low-frequency component at the edge of radiation plates to improve the low frequency character of the antenna. In this paper, an ultra-wideband TEM horn antenna improved with the proposed structure is presented. Finally, the comparison of the improved antenna and the original one is exhibited, which shows that the impedance bandwidth is 12.5% wider with the low-cutoff frequency decreasing to 0.1 GHz and simultaneously the port-feed efficiency is increasing by 10%. The results confirm the validity of the proposed design to optimize the low frequency character. The feasibility is also validated in the end by analyzing the impedance and the magnitude distribution of currents in different frequencies, respectively.
2016, 38(11): 2960-2964.
doi: 10.11999/JEIT160608
Abstract:
To improve the environmental adaptation of the package of electric field micro sensor, this article introducs a new electrode package. Different from placing the sensing chip and package inside the measured environment, a package electrode, which is the only part exposed outside, is invented. In this way, the package shell is effectively protected from various environmental extremes. Based on this structure, new atmospheric electric field micro sensors are introduced on the ground and for sounding. The FEM and experiments show that the new sensors measure electric field accurately and stably under high humidity and low temperature environment.
To improve the environmental adaptation of the package of electric field micro sensor, this article introducs a new electrode package. Different from placing the sensing chip and package inside the measured environment, a package electrode, which is the only part exposed outside, is invented. In this way, the package shell is effectively protected from various environmental extremes. Based on this structure, new atmospheric electric field micro sensors are introduced on the ground and for sounding. The FEM and experiments show that the new sensors measure electric field accurately and stably under high humidity and low temperature environment.
2016, 38(11): 2965-2971.
doi: 10.11999/JEIT160035
Abstract:
As the core component of space traveling wave tubes, the cathode-heater assembly is required to be stable, reliable, long life and low power consumption. In this paper, an estimation method of thermal contact resistance is proposed, and the thermal characteristics of cathode-heater assembly structure are simulated. Meanwhile, thermal experiment is designed and undertaken, and the whole temperature distribution of cathode-heater assembly structure under a variety of heating power is obtained for the first time. Furthermore, the thermal boundary and excitation of cathode-heater assembly structure is modified, and the values of thermal contact resistances are obtained by interactive method. Finally, a high reliable thermal model of cathode-heater assembly structure is obtained. It is revealed that cathode temperature calculating accuracy is within 5%, and the calculating error of whole structure is less than72 C.
As the core component of space traveling wave tubes, the cathode-heater assembly is required to be stable, reliable, long life and low power consumption. In this paper, an estimation method of thermal contact resistance is proposed, and the thermal characteristics of cathode-heater assembly structure are simulated. Meanwhile, thermal experiment is designed and undertaken, and the whole temperature distribution of cathode-heater assembly structure under a variety of heating power is obtained for the first time. Furthermore, the thermal boundary and excitation of cathode-heater assembly structure is modified, and the values of thermal contact resistances are obtained by interactive method. Finally, a high reliable thermal model of cathode-heater assembly structure is obtained. It is revealed that cathode temperature calculating accuracy is within 5%, and the calculating error of whole structure is less than72 C.
2016, 38(11): 2972-2976.
doi: 10.11999/JEIT160031
Abstract:
A novel recognition algorithm for Low Probability of Intercept (LPI) radar signal based on deep learning of radar signals Bispectra Diagonal Slice (BDS) is proposed in this paper. Firstly, a Deep Belief Network (DBN) model is established on stacked Restricted Boltzmann Machines (RBM), then the model is used for layer-by-layer unsupervised greedy learning of radar signals BDS. Secondly, a Back Propagation (BP) algorithm is applied to fine tune parameters of DBN model with a supervised way according to learning error. Finally, the BDS-DBN model is constructed to classify and recognize unknown LPI signals. The theoretical analysis and the simulation results show that, the average recognition accuracy of the proposed algorithm for Frequency Modulation Continuous Wave (FMCW), Frank, Costas and FSK/PSK signals can reach 93.4% or ever higher while the SNR is better than 8 dB, which is better than that of Principal Component Analysis-Support Vector Machine (PCA-SVM) algorithm and Principal Component Analysis-Linear Discriminant Analysis (PCA-LDA) algorithm.
A novel recognition algorithm for Low Probability of Intercept (LPI) radar signal based on deep learning of radar signals Bispectra Diagonal Slice (BDS) is proposed in this paper. Firstly, a Deep Belief Network (DBN) model is established on stacked Restricted Boltzmann Machines (RBM), then the model is used for layer-by-layer unsupervised greedy learning of radar signals BDS. Secondly, a Back Propagation (BP) algorithm is applied to fine tune parameters of DBN model with a supervised way according to learning error. Finally, the BDS-DBN model is constructed to classify and recognize unknown LPI signals. The theoretical analysis and the simulation results show that, the average recognition accuracy of the proposed algorithm for Frequency Modulation Continuous Wave (FMCW), Frank, Costas and FSK/PSK signals can reach 93.4% or ever higher while the SNR is better than 8 dB, which is better than that of Principal Component Analysis-Support Vector Machine (PCA-SVM) algorithm and Principal Component Analysis-Linear Discriminant Analysis (PCA-LDA) algorithm.
2016, 38(11): 2977-2981.
doi: 10.11999/JEIT160089
Abstract:
The Theory of Characteristic Modes (TCM) provides the antenna designer with physical insight regarding the antenna operating principles. Based on TCM, this paper analyzes in detail the characteristic modes properties of the several transfigurations of metal ring in a cylinder cavity. These properties give information about the radiation ability of every structure. It is very helpful to choose the very right structure as the desired antenna. This paper also gives an effective feed for the combined Inverse-L mode. The manufactured antenna is measured. The measured results show that the band width is 1.8% for VSWR2, the diameter is 0.260, the height is 0.080, and the measured results agree well with the simulated ones. It is verified that the design method based on TCM is very effective.
The Theory of Characteristic Modes (TCM) provides the antenna designer with physical insight regarding the antenna operating principles. Based on TCM, this paper analyzes in detail the characteristic modes properties of the several transfigurations of metal ring in a cylinder cavity. These properties give information about the radiation ability of every structure. It is very helpful to choose the very right structure as the desired antenna. This paper also gives an effective feed for the combined Inverse-L mode. The manufactured antenna is measured. The measured results show that the band width is 1.8% for VSWR2, the diameter is 0.260, the height is 0.080, and the measured results agree well with the simulated ones. It is verified that the design method based on TCM is very effective.
2016, 38(11): 2982-2986.
doi: 10.11999/JEIT160140
Abstract:
In order to alleviate the influence of network congestion in space delay/disruption tolerant networks, a QoS based congestion control algorithm is proposed in this paper. The algorithm consists of contact congestion forecasting scheme and QoS based data forwarding scheme. The contacts are divided into different congestion levels according to their residual available capacity and nodes storage resource. The congestion level of forwarding path calculated by routing is decided by the highest congestion level of contacts consisted in the path and different priority data will be forwarded according to the path congestion level. Experiment shows that QoS based algorithm could improve the transmission rate of data with lower priority and reduce the delivery delay of highest priority data when the node storage space is insufficient.
In order to alleviate the influence of network congestion in space delay/disruption tolerant networks, a QoS based congestion control algorithm is proposed in this paper. The algorithm consists of contact congestion forecasting scheme and QoS based data forwarding scheme. The contacts are divided into different congestion levels according to their residual available capacity and nodes storage resource. The congestion level of forwarding path calculated by routing is decided by the highest congestion level of contacts consisted in the path and different priority data will be forwarded according to the path congestion level. Experiment shows that QoS based algorithm could improve the transmission rate of data with lower priority and reduce the delivery delay of highest priority data when the node storage space is insufficient.
