Email alert
2018 Vol. 40, No. 7
Display Method:
2018, 40(7): 1525-1531.
doi: 10.11999/JEIT170990
Abstract:
With the improvement of imaging resolution and on-orbit mobility of earth observation satellites, the imaging geometric quality is more apparently influenced by the attitude’s high-frequency jittering of satellite platform. The traditional time-division imaging data based jitter detection and compensation methods have many drawbacks, which include large amount of calculation and high degree of error interference in dense matching, and it is unable to decompose the jitter quantity in each rotation angle direction. This paper takes the high-frequency angular displacement equipment which is carried by China’s remote sensing optical satellite for example, studies on the direct jitter detection method and the image geometric quality compensation method based on high-frequency attitude measurement angular displacement data, which include the windowed FIR filter pre-processing of angular displacement data, the phase distribution analysis on time-dependent jitter curve in pitch, roll and yaw directions, as well as image direct positioning compensation based on angular displacement data. The high-frequency jitter compensation is applied to attitude recovery and geometric rectification based on strict imaging geometric model.The experimental results of China’ remote sensing satellite images in Beijing area illustrate that the methods proposed in this paper can significantly improve the accuracy and reliability of the high- frequency jitter detection, and can effectively improve the internal geometric quality of satellite image after jitter compensation. For example, the length deformation accuracy can be improved by 0.5 pixel.
With the improvement of imaging resolution and on-orbit mobility of earth observation satellites, the imaging geometric quality is more apparently influenced by the attitude’s high-frequency jittering of satellite platform. The traditional time-division imaging data based jitter detection and compensation methods have many drawbacks, which include large amount of calculation and high degree of error interference in dense matching, and it is unable to decompose the jitter quantity in each rotation angle direction. This paper takes the high-frequency angular displacement equipment which is carried by China’s remote sensing optical satellite for example, studies on the direct jitter detection method and the image geometric quality compensation method based on high-frequency attitude measurement angular displacement data, which include the windowed FIR filter pre-processing of angular displacement data, the phase distribution analysis on time-dependent jitter curve in pitch, roll and yaw directions, as well as image direct positioning compensation based on angular displacement data. The high-frequency jitter compensation is applied to attitude recovery and geometric rectification based on strict imaging geometric model.The experimental results of China’ remote sensing satellite images in Beijing area illustrate that the methods proposed in this paper can significantly improve the accuracy and reliability of the high- frequency jitter detection, and can effectively improve the internal geometric quality of satellite image after jitter compensation. For example, the length deformation accuracy can be improved by 0.5 pixel.
2018, 40(7): 1532-1538.
doi: 10.11999/JEIT170961
Abstract:
Unmanned Aerial Vehicle (UAV) 3D path planning is the most complex and important part of mission planning. Considering at the problem that the problem of 3D path planning can not be solved by the original algorithm perfectly, so firstly the chaotic adjustment factor and anti-regulation factor are introduced into the behavior of ant and ant lion respectively, which improves the exploration and the exploitation of algorithm. Then, in order to reduce search space ,so terrain and constraints are full used on the basis of the establishment of 3D environment model. Lastly, the improved algorithm is applied to the 3D path planning, which is compared with the original algorithm, and online local re-planning is implemented. Simulation results demonstrate the feasibility and superiority of the improved method.
Unmanned Aerial Vehicle (UAV) 3D path planning is the most complex and important part of mission planning. Considering at the problem that the problem of 3D path planning can not be solved by the original algorithm perfectly, so firstly the chaotic adjustment factor and anti-regulation factor are introduced into the behavior of ant and ant lion respectively, which improves the exploration and the exploitation of algorithm. Then, in order to reduce search space ,so terrain and constraints are full used on the basis of the establishment of 3D environment model. Lastly, the improved algorithm is applied to the 3D path planning, which is compared with the original algorithm, and online local re-planning is implemented. Simulation results demonstrate the feasibility and superiority of the improved method.
2018, 40(7): 1539-1545.
doi: 10.11999/JEIT170954
Abstract:
L-band passive microwave remote sensing is an effective method for detection of soil moisture and ocean salinity. However, spectrum pollution from Global Positioning System (GPS), radar and electromagnetic radiation from some commercial electronic products can interfere with the detection of microwave radiometers, it results in a certain deviation to the ground observation, and reduces the retrieval accuracy of the surface parameters. Pulsed noise interference is simulated by experiment, its transmission characteristics in L-band (total power receiver) microwave radiometer system is observed, the correlation between the characteristics of the output signal and the radiometer parameters (integrating time and sensitivity) is analyzed, its digital characteristic parameters are obtained. Combining the method of Asynchronous Pulse Blanking (APB), a new AutoCorrelation Detection (ACD) algorithm is proposed, which can detect periodic pulsed radiation interference effectively. In the case that microwave radiometer system integrating time is 1 ms, it can detect the noise interference of 1.5 K, which meets the requirement of precision of surface parameters retrieved by satellite remote sensing.
L-band passive microwave remote sensing is an effective method for detection of soil moisture and ocean salinity. However, spectrum pollution from Global Positioning System (GPS), radar and electromagnetic radiation from some commercial electronic products can interfere with the detection of microwave radiometers, it results in a certain deviation to the ground observation, and reduces the retrieval accuracy of the surface parameters. Pulsed noise interference is simulated by experiment, its transmission characteristics in L-band (total power receiver) microwave radiometer system is observed, the correlation between the characteristics of the output signal and the radiometer parameters (integrating time and sensitivity) is analyzed, its digital characteristic parameters are obtained. Combining the method of Asynchronous Pulse Blanking (APB), a new AutoCorrelation Detection (ACD) algorithm is proposed, which can detect periodic pulsed radiation interference effectively. In the case that microwave radiometer system integrating time is 1 ms, it can detect the noise interference of 1.5 K, which meets the requirement of precision of surface parameters retrieved by satellite remote sensing.
2018, 40(7): 1546-1552.
doi: 10.11999/JEIT170896
Abstract:
When ship target is monitored by the geostationary optical satellite, the positioning error is large due to the long distance between the target and the satellite, which affects the accuracy of the follow-up target tracking. As the monitoring area is mainly the ocean, it may not be possible to find the Ground Control Point (GCP) for coordinate correction. In order to improve the positioning accuracy of the geostationary optical satellite for ship without GCP, and to realize the fusion of multi-source data, a novel target point association and error correction with optical satellite in geostationary orbit and ship Automatic Identification System (AIS) is proposed. By means of the Rational Polynomial Coefficient (RPC) model, AIS coordinates are transformed into image coordinates. The Iterative Closest Point (ICP) and Global Nearest Neighbor (GNN) algorithm are combined and used for data association. Then, the error is corrected using the point pair of association. Experimental results using GF-4 images and AIS data verify the feasibility of the proposed method and show that the association algorithm has a high correlation rate, and the average positioning accuracy after error correction is improved greatly compared with the positioning accuracy before correction.
When ship target is monitored by the geostationary optical satellite, the positioning error is large due to the long distance between the target and the satellite, which affects the accuracy of the follow-up target tracking. As the monitoring area is mainly the ocean, it may not be possible to find the Ground Control Point (GCP) for coordinate correction. In order to improve the positioning accuracy of the geostationary optical satellite for ship without GCP, and to realize the fusion of multi-source data, a novel target point association and error correction with optical satellite in geostationary orbit and ship Automatic Identification System (AIS) is proposed. By means of the Rational Polynomial Coefficient (RPC) model, AIS coordinates are transformed into image coordinates. The Iterative Closest Point (ICP) and Global Nearest Neighbor (GNN) algorithm are combined and used for data association. Then, the error is corrected using the point pair of association. Experimental results using GF-4 images and AIS data verify the feasibility of the proposed method and show that the association algorithm has a high correlation rate, and the average positioning accuracy after error correction is improved greatly compared with the positioning accuracy before correction.
2018, 40(7): 1553-1560.
doi: 10.11999/JEIT170931
Abstract:
Pulse radar signal is widely used in Inverse Synthetic Aperture Radar (ISAR) imaging. However, because the propagation distance obtained by the pulse width is larger than the size of anechoic chamber, target echo returns before the pulse signal is fully transmitted. As a result, the transmitted and reflected signals are coupled and difficult to be separated for ISAR imaging in Radio Frequency Simulation (RFS). To solve this problem, the equivalent simulation method is proposed for pulse radar ISAR imaging based on Interrupted Transmitting and Receiving (ITR). As the ITR echo is piecewise sparse in time domain, the ISAR image can be reconstructed based on Compressive Sensing (CS). Simulation and real data demonstrate that the ISAR image obtained by the proposed method is consistent with that obtained by the complete pulse radar signal. Therefore, the proposed method is effective for ISAR imaging in RFS.
Pulse radar signal is widely used in Inverse Synthetic Aperture Radar (ISAR) imaging. However, because the propagation distance obtained by the pulse width is larger than the size of anechoic chamber, target echo returns before the pulse signal is fully transmitted. As a result, the transmitted and reflected signals are coupled and difficult to be separated for ISAR imaging in Radio Frequency Simulation (RFS). To solve this problem, the equivalent simulation method is proposed for pulse radar ISAR imaging based on Interrupted Transmitting and Receiving (ITR). As the ITR echo is piecewise sparse in time domain, the ISAR image can be reconstructed based on Compressive Sensing (CS). Simulation and real data demonstrate that the ISAR image obtained by the proposed method is consistent with that obtained by the complete pulse radar signal. Therefore, the proposed method is effective for ISAR imaging in RFS.
2018, 40(7): 1561-1566.
doi: 10.11999/JEIT170996
Abstract:
In multichannel High-Resolution and Wide-Swath (HRWS) Synthetic Aperture Radar (SAR) systems, channel amplitude and phase mismatches would degrade the performance of azimuth ambiguity suppression as well as range sampling mismatches. To address this problem, a robust estimation method based on range spectrum analysis is proposed in this letter. The method includes two steps: firstly, by exploiting the interferometric phases between the range spectrums of adjacent channels, the proposed method can robustly estimate range sampling mismatches by the combination of phase unwrapping and a weighted least squares fitting. Secondly, based on the theory of spatial cross correlation, an accurate Doppler centroid and phase mismatches could be obtained from constant phases between adjacent channels. Compared with traditional methods, the proposed method overcomes the effect of phase wraps and jumps on polynomial coefficient estimation. It not only improves the robustness of parameter estimation but also can simultaneously obtain the Doppler centroid and phase mismatches. Experimental results based on airborne real data and simulated data validate its effectiveness of the proposed method.
In multichannel High-Resolution and Wide-Swath (HRWS) Synthetic Aperture Radar (SAR) systems, channel amplitude and phase mismatches would degrade the performance of azimuth ambiguity suppression as well as range sampling mismatches. To address this problem, a robust estimation method based on range spectrum analysis is proposed in this letter. The method includes two steps: firstly, by exploiting the interferometric phases between the range spectrums of adjacent channels, the proposed method can robustly estimate range sampling mismatches by the combination of phase unwrapping and a weighted least squares fitting. Secondly, based on the theory of spatial cross correlation, an accurate Doppler centroid and phase mismatches could be obtained from constant phases between adjacent channels. Compared with traditional methods, the proposed method overcomes the effect of phase wraps and jumps on polynomial coefficient estimation. It not only improves the robustness of parameter estimation but also can simultaneously obtain the Doppler centroid and phase mismatches. Experimental results based on airborne real data and simulated data validate its effectiveness of the proposed method.
2018, 40(7): 1567-1574.
doi: 10.11999/JEIT170913
Abstract:
In modern warfare, the radar system is developing rapidly. To recognize complex modulation mode of radar signal and hybrid pulse repetition interval radar, this paper proposes a sorting method based on multi station acquired pulse time-difference parameter combined with other pulse description words, taking advantage of the multi-station TDOA from the same emitter is similar to sort emitter signal pulse, and finally got the recognition result with the Multi-Layer Percetron (MLP) neural network. Traditional Pulse Repetition Interval (PRI) estimation algorithms estimate complex pulse interval modulation invalidly. In this paper, to solve this problem pulse time-difference parameter and other pulse description words are used. The feature vector of de-interlace pulse sequence is acquired and the result of pulse interval modulation type recognition is obtained with the trained MLP neutral network. Through experimental simulation, the correct recognition probability of the complex pulse interval modulation method is more than 90% in the case of the pulse loss rate is not more than 20%.
In modern warfare, the radar system is developing rapidly. To recognize complex modulation mode of radar signal and hybrid pulse repetition interval radar, this paper proposes a sorting method based on multi station acquired pulse time-difference parameter combined with other pulse description words, taking advantage of the multi-station TDOA from the same emitter is similar to sort emitter signal pulse, and finally got the recognition result with the Multi-Layer Percetron (MLP) neural network. Traditional Pulse Repetition Interval (PRI) estimation algorithms estimate complex pulse interval modulation invalidly. In this paper, to solve this problem pulse time-difference parameter and other pulse description words are used. The feature vector of de-interlace pulse sequence is acquired and the result of pulse interval modulation type recognition is obtained with the trained MLP neutral network. Through experimental simulation, the correct recognition probability of the complex pulse interval modulation method is more than 90% in the case of the pulse loss rate is not more than 20%.
2018, 40(7): 1575-1582.
doi: 10.11999/JEIT171068
Abstract:
Considering the problem of large squint synthetic aperture sonar imaging, the analytical expression of the wavenumber spectrum is analyzed in detail in the radial and azimuth wavenumber fields under the wide-band high-squint conditions. The spectrum winding and shrink in the distance wavenumber fields after the Stolt interpolation are pointed out, and the reduced relative distance between the target in the imaging result is also indicated, then the Stolt interpolation method for distance wavenumber spectrum winding is given. The concept of range wavenumber scaling factor is proposed, the method of compensating the scaling factor and the spectrum winding in the distance space are given. Finally, the problem of range scaling caused by Stolt interpolation under large oblique angle is solved by compensating the distance variable in distance space. Point object simulation data and simulated echo data processing verify the correctness and validity of the proposed method.
Considering the problem of large squint synthetic aperture sonar imaging, the analytical expression of the wavenumber spectrum is analyzed in detail in the radial and azimuth wavenumber fields under the wide-band high-squint conditions. The spectrum winding and shrink in the distance wavenumber fields after the Stolt interpolation are pointed out, and the reduced relative distance between the target in the imaging result is also indicated, then the Stolt interpolation method for distance wavenumber spectrum winding is given. The concept of range wavenumber scaling factor is proposed, the method of compensating the scaling factor and the spectrum winding in the distance space are given. Finally, the problem of range scaling caused by Stolt interpolation under large oblique angle is solved by compensating the distance variable in distance space. Point object simulation data and simulated echo data processing verify the correctness and validity of the proposed method.
2018, 40(7): 1583-1590.
doi: 10.11999/JEIT171025
Abstract:
Research on target tracking with glint noise is important to improve detection performance of sensor, in which the glint noise’s unknown distribution and non-stationary property puzzle researchers for a long time. In order to solve this problem, the tracking theoretical framework of variational Bayesian parameter learning with glint noise is firstly introduced. Then, a novel algorithm called Variational Bayesian-Interacting Multiple Model (VB-IMM) is proposed to estimate the system states as well as the unknown glint noise’s distribution. The proposed algorithm designs a bank of tracking filters in parallel with different measurement noise. Moreover, the algorithm utilizes variational Bayesian method to learn distribution parameters of the glint noise online and feed these parameters back to the tracking filters to revise the filters. In order to validate the performance of this algorithm, comparative experiments are carried out from two aspects of tracking accuracy and computational complexity. Simulation results verify good performance of tracking error and low computational complexity of the proposed algorithm.
Research on target tracking with glint noise is important to improve detection performance of sensor, in which the glint noise’s unknown distribution and non-stationary property puzzle researchers for a long time. In order to solve this problem, the tracking theoretical framework of variational Bayesian parameter learning with glint noise is firstly introduced. Then, a novel algorithm called Variational Bayesian-Interacting Multiple Model (VB-IMM) is proposed to estimate the system states as well as the unknown glint noise’s distribution. The proposed algorithm designs a bank of tracking filters in parallel with different measurement noise. Moreover, the algorithm utilizes variational Bayesian method to learn distribution parameters of the glint noise online and feed these parameters back to the tracking filters to revise the filters. In order to validate the performance of this algorithm, comparative experiments are carried out from two aspects of tracking accuracy and computational complexity. Simulation results verify good performance of tracking error and low computational complexity of the proposed algorithm.
2018, 40(7): 1591-1597.
doi: 10.11999/JEIT170979
Abstract:
To solve the nonlinear equation problems of Time-Difference-Of-Arrival (TDOA) passive location, a new swarm intelligence optimization algorithm called Salp-Swarm-Algorithm (SSA) is used. Firstly, a new renewal model of salps is proposed to balance exploration and exploitation properly during iteration in SSA. SSA not only ensures the wholeness of searching and the diversity of individuals, but also improves the problem that other intelligent optimization algorithms fall into local optima easily. Besides, there are few parameters to be adjusted, therefor, the computation speed is obviously improved. Moreover, the convergence performance of the proposed algorithm is very stable and the accuracy of location is higher. Simulation results show that the proposed algorithm can converge to the position of emitters fast and stably in 3D TDOA location. Comparing with Particle-Swarm- Optimization (PSO) and Improved-Particle-Swarm-Optimization (IPSO), the proposed algorithm has lower mean square error.
To solve the nonlinear equation problems of Time-Difference-Of-Arrival (TDOA) passive location, a new swarm intelligence optimization algorithm called Salp-Swarm-Algorithm (SSA) is used. Firstly, a new renewal model of salps is proposed to balance exploration and exploitation properly during iteration in SSA. SSA not only ensures the wholeness of searching and the diversity of individuals, but also improves the problem that other intelligent optimization algorithms fall into local optima easily. Besides, there are few parameters to be adjusted, therefor, the computation speed is obviously improved. Moreover, the convergence performance of the proposed algorithm is very stable and the accuracy of location is higher. Simulation results show that the proposed algorithm can converge to the position of emitters fast and stably in 3D TDOA location. Comparing with Particle-Swarm- Optimization (PSO) and Improved-Particle-Swarm-Optimization (IPSO), the proposed algorithm has lower mean square error.
2018, 40(7): 1598-1603.
doi: 10.11999/JEIT170906
Abstract:
Existing methods for symbol rate estimation of phase coded signals require amounts of sensing data, and are of high computational complexity. This paper analyzes the structure characteristics of BPSK signals, which are employed as the prior information for signal compressing and dimensionality reduction. The sensing matrix can be split into sine and cosine component, combined with the Fourier transform parity. According to the fact that the real and imaginary components of a complex value share the same support set, the symbol rate estimation can be obtained, using unilateral spectral of the delay-product vector reconstructed by multi-task Bayesian compressive sensing. Theoretical analysis and simulation results show that compared with other parameter estimation algorithms, the proposed method can reduce the measurements and significantly improve the real-time ability, while keeping the high reconstruction accuracy.
Existing methods for symbol rate estimation of phase coded signals require amounts of sensing data, and are of high computational complexity. This paper analyzes the structure characteristics of BPSK signals, which are employed as the prior information for signal compressing and dimensionality reduction. The sensing matrix can be split into sine and cosine component, combined with the Fourier transform parity. According to the fact that the real and imaginary components of a complex value share the same support set, the symbol rate estimation can be obtained, using unilateral spectral of the delay-product vector reconstructed by multi-task Bayesian compressive sensing. Theoretical analysis and simulation results show that compared with other parameter estimation algorithms, the proposed method can reduce the measurements and significantly improve the real-time ability, while keeping the high reconstruction accuracy.
2018, 40(7): 1604-1611.
doi: 10.11999/JEIT170951
Abstract:
In order to improve DOA estimation accuracy of co-prime array while the number of snapshots is small, a novel fast Sparse Bayesian Learning (SBL) algorithm using Bessel priors is proposed. Focusing on the multi-snapshots complex output data of coprime array, a multiple measurement vectors hierarchical model based on Bessel priors is firstly built. Then the log-likelihood function of model hyperparameters is derived, and the iterative formulas of hyperparameters are derived based on the criterion of maximum likelihood estimation. Finally, a fast implementation scheme is developed in order to improve the computation efficiency. Simulation experiments show that the proposed algorithm is independent on prior information. Under the condition of small number of snapshots, higher DOA estimation accuracy and resolution of uncorrelated and correlated signals can be achieved with proper computational efficiency. Further more, the necessity between virtual array extension and DOA estimation freedom of co-prime array is explored, which provides reference to DOA estimation for co-prime array under array perturbation conditions.
In order to improve DOA estimation accuracy of co-prime array while the number of snapshots is small, a novel fast Sparse Bayesian Learning (SBL) algorithm using Bessel priors is proposed. Focusing on the multi-snapshots complex output data of coprime array, a multiple measurement vectors hierarchical model based on Bessel priors is firstly built. Then the log-likelihood function of model hyperparameters is derived, and the iterative formulas of hyperparameters are derived based on the criterion of maximum likelihood estimation. Finally, a fast implementation scheme is developed in order to improve the computation efficiency. Simulation experiments show that the proposed algorithm is independent on prior information. Under the condition of small number of snapshots, higher DOA estimation accuracy and resolution of uncorrelated and correlated signals can be achieved with proper computational efficiency. Further more, the necessity between virtual array extension and DOA estimation freedom of co-prime array is explored, which provides reference to DOA estimation for co-prime array under array perturbation conditions.
2018, 40(7): 1612-1618.
doi: 10.11999/JEIT171014
Abstract:
In recent years, Teager energy operator is proposed as a kind of nonlinear method characterized with tracking a time-varying signal. The operator is combined with empirical mode decomposition, and a new method of voice activity detection is proposed to find the best voice start point and end point. Empirical Mode Decomposition (EMD) is further exploited and some valid choice conditions are constructed to choose the valid intrinsic mode functions. Thus, the method is able to deal with the voice with noise. Also, the character of the single mode of empirical mode decomposition meets the demand of single frequency component required by Teager Energy Operator (TEO). At last, Hilbert transform is added to solve the inherent problem of the mode mixing due to empirical mode decomposition. Based on the above consideration, the proposed method can identify the unvoiced sound with noise, which is better than the direct TEO and double threshold method. Experiments show the validity of the proposed method.
In recent years, Teager energy operator is proposed as a kind of nonlinear method characterized with tracking a time-varying signal. The operator is combined with empirical mode decomposition, and a new method of voice activity detection is proposed to find the best voice start point and end point. Empirical Mode Decomposition (EMD) is further exploited and some valid choice conditions are constructed to choose the valid intrinsic mode functions. Thus, the method is able to deal with the voice with noise. Also, the character of the single mode of empirical mode decomposition meets the demand of single frequency component required by Teager Energy Operator (TEO). At last, Hilbert transform is added to solve the inherent problem of the mode mixing due to empirical mode decomposition. Based on the above consideration, the proposed method can identify the unvoiced sound with noise, which is better than the direct TEO and double threshold method. Experiments show the validity of the proposed method.
2018, 40(7): 1619-1626.
doi: 10.11999/JEIT170939
Abstract:
In traditional sparse representation based visual tracking, particle sampling is first achieved by particle filter method. Then the particle observations are represented by intensity feature. Finally, the visual tracking is achieved by the intensity feature based sparse representation model. Different from traditional sparse representation model, a canonical correlation analysis based sparse representation model is proposed in this paper. The proposed model first uses two kinds of features to represent the particle observations, then, the projections of particle observations are used to build the sparse representation model. The advantage of the proposed model lies in that it can give a proper multi-feature fusing through canonical correlation analysis, which explores the relation between two features in a latent common subspace.
In traditional sparse representation based visual tracking, particle sampling is first achieved by particle filter method. Then the particle observations are represented by intensity feature. Finally, the visual tracking is achieved by the intensity feature based sparse representation model. Different from traditional sparse representation model, a canonical correlation analysis based sparse representation model is proposed in this paper. The proposed model first uses two kinds of features to represent the particle observations, then, the projections of particle observations are used to build the sparse representation model. The advantage of the proposed model lies in that it can give a proper multi-feature fusing through canonical correlation analysis, which explores the relation between two features in a latent common subspace.
2018, 40(7): 1627-1634.
doi: 10.11999/JEIJ170995
Abstract:
The high realism simulation interaction of flower plants is an important direction of virtual plant visualization. More and more applications are presented by the Virtual Reality (VR) headset device with the popularity of virtual reality technology. The VR system requires a highly realistic immersive picture for which generic plant modeling and graphics engine rendering capabilities are no longer sufficient. In this paper, a physical rendering algorithm is proposed based on Bidirectional Scattering Distribution Function (BSDF) to realistic flower plants by analyzing the principle of illumination and combining the Physics-Based Shading (PBS) technology. Potted flora in light mode is simulated by ShaderLab and the fusion algorithm is optimized. The image is distorted by lens matching rendering technology to make the virtual scene closer to the real human eye vision and reduce the user vertigo when user wearing the VR headset when using the VR helmet equipment HTC Vive. Finally a helmet VR floral plant simulation system is designed and a realistic immersive scene is realized.
The high realism simulation interaction of flower plants is an important direction of virtual plant visualization. More and more applications are presented by the Virtual Reality (VR) headset device with the popularity of virtual reality technology. The VR system requires a highly realistic immersive picture for which generic plant modeling and graphics engine rendering capabilities are no longer sufficient. In this paper, a physical rendering algorithm is proposed based on Bidirectional Scattering Distribution Function (BSDF) to realistic flower plants by analyzing the principle of illumination and combining the Physics-Based Shading (PBS) technology. Potted flora in light mode is simulated by ShaderLab and the fusion algorithm is optimized. The image is distorted by lens matching rendering technology to make the virtual scene closer to the real human eye vision and reduce the user vertigo when user wearing the VR headset when using the VR helmet equipment HTC Vive. Finally a helmet VR floral plant simulation system is designed and a realistic immersive scene is realized.
2018, 40(7): 1635-1642.
doi: 10.11999/JEIT171011
Abstract:
Cross-modal speaker tagging aims to learn the latent relationship between different biometrics for mutual annotation, which can potentially be utilized in various human-computer interactions. In order to solve the “semantic gap” between the face and audio modalities, this paper presents an efficient supervised joint correspondence auto-encoder to link the face and audio counterpart, where by the speaker can be crosswise tagged. First, Convolutional Neural Network (CNN) and Deep Belief Network (DBN) are used to extract the discriminative features of the face and the audio samples respectively. Then, a supervised neural network model associated with softmax regression is embedded into a joint auto-encoder model, which can discriminatively preserving the inter-modal and intra-modal similarities. Accordingly, three different kinds of supervised joint correspondence auto-encoder models are presented to correlate the semantic relationships between the face and the audio counterparts, and the speaker can be crosswise annotated efficiently. The experimental results show that the proposed supervised joint auto-encoder is able to perform cross-modal speaker tagging with outstanding performance, and demonstrate the robustness to facial posture variations and sample diversities.
Cross-modal speaker tagging aims to learn the latent relationship between different biometrics for mutual annotation, which can potentially be utilized in various human-computer interactions. In order to solve the “semantic gap” between the face and audio modalities, this paper presents an efficient supervised joint correspondence auto-encoder to link the face and audio counterpart, where by the speaker can be crosswise tagged. First, Convolutional Neural Network (CNN) and Deep Belief Network (DBN) are used to extract the discriminative features of the face and the audio samples respectively. Then, a supervised neural network model associated with softmax regression is embedded into a joint auto-encoder model, which can discriminatively preserving the inter-modal and intra-modal similarities. Accordingly, three different kinds of supervised joint correspondence auto-encoder models are presented to correlate the semantic relationships between the face and the audio counterparts, and the speaker can be crosswise annotated efficiently. The experimental results show that the proposed supervised joint auto-encoder is able to perform cross-modal speaker tagging with outstanding performance, and demonstrate the robustness to facial posture variations and sample diversities.
2018, 40(7): 1643-1651.
doi: 10.11999/JEIT170984
Abstract:
The key to the success of an ensemble system are the diversity and the average accuracy of base classifiers. The increase of diversity among base classifiers will lead to the decrease of the average accuracy, and vice versa. So there exists a tradeoff between the diversity and the average accuracy, which makes the ensemble perform the best with respect to ensemble pruning. To find the tradeoff, Improved Binary Glowworm Swarm Optimization combined with Complementarity measure for Ensemble Pruning (IBGSOCEP) is proposed. Firstly, an initial pool of classifiers is constructed through training independently some base classifiers using bootstrap sampling. Secondly, the classifiers in the initial pool are pre-pruned using complementarity measure. Thirdly, Improved Binary Glowworm Swarm Optimization (IBGSO) is proposed by improving moving way, searching processes of glowworm, introducing re-initialization, and leaping behaviors. Finally, the optimal sub-ensemble is achieved from the base classifiers after pre-pruning using IBGSO. Experimental results on 5 UCI datasets demonstrate that IBGSODSEN can achieve better results than other approaches with less number of base classifiers, and that its effectiveness and significance.
The key to the success of an ensemble system are the diversity and the average accuracy of base classifiers. The increase of diversity among base classifiers will lead to the decrease of the average accuracy, and vice versa. So there exists a tradeoff between the diversity and the average accuracy, which makes the ensemble perform the best with respect to ensemble pruning. To find the tradeoff, Improved Binary Glowworm Swarm Optimization combined with Complementarity measure for Ensemble Pruning (IBGSOCEP) is proposed. Firstly, an initial pool of classifiers is constructed through training independently some base classifiers using bootstrap sampling. Secondly, the classifiers in the initial pool are pre-pruned using complementarity measure. Thirdly, Improved Binary Glowworm Swarm Optimization (IBGSO) is proposed by improving moving way, searching processes of glowworm, introducing re-initialization, and leaping behaviors. Finally, the optimal sub-ensemble is achieved from the base classifiers after pre-pruning using IBGSO. Experimental results on 5 UCI datasets demonstrate that IBGSODSEN can achieve better results than other approaches with less number of base classifiers, and that its effectiveness and significance.
2018, 40(7): 1652-1660.
doi: 10.11999/JEIT170863
Abstract:
The existing Ciphertext-Policy Attribute-Based Encryption (CP-ABE) schemes from lattices are inefficient while they are performed in matrix operation, and these Key-Policy Attribute-Based Encryption (KP-ABE) schemes from ideal lattices with higher efficiency are inadaptable to most practical application scenarios. To solve these problems, the new scheme generates master keys and secret keys by the algorithms based on ideal lattices and the whole scheme is computed over a polynomial ring, thus its efficiency of encryption and decryption can be greatly improved. The ciphertexts associated with access structure are successfully generated by adding some virtual attributes to the original attribute set. Meanwhile, the authorized user can build a subset based on these virtual attributes for decrypting the scheme correctly. And the secret keys are generated by a single trapdoor matrix, which reduces the number of public parameters and master keys effectively. Finally, an efficient CP-ABE scheme for flexible threshold access structures on ideal lattices is proposed, and its security is reduced to decisional Learning With Errors over Ring (R-LWE) assumption against chosen plaintext attack in the selective security model. Comparative analysis of similar schemes shows that the new scheme has less public parameters and higher efficiency, and gets better adaptability to the practical application scenarios.
The existing Ciphertext-Policy Attribute-Based Encryption (CP-ABE) schemes from lattices are inefficient while they are performed in matrix operation, and these Key-Policy Attribute-Based Encryption (KP-ABE) schemes from ideal lattices with higher efficiency are inadaptable to most practical application scenarios. To solve these problems, the new scheme generates master keys and secret keys by the algorithms based on ideal lattices and the whole scheme is computed over a polynomial ring, thus its efficiency of encryption and decryption can be greatly improved. The ciphertexts associated with access structure are successfully generated by adding some virtual attributes to the original attribute set. Meanwhile, the authorized user can build a subset based on these virtual attributes for decrypting the scheme correctly. And the secret keys are generated by a single trapdoor matrix, which reduces the number of public parameters and master keys effectively. Finally, an efficient CP-ABE scheme for flexible threshold access structures on ideal lattices is proposed, and its security is reduced to decisional Learning With Errors over Ring (R-LWE) assumption against chosen plaintext attack in the selective security model. Comparative analysis of similar schemes shows that the new scheme has less public parameters and higher efficiency, and gets better adaptability to the practical application scenarios.
2018, 40(7): 1661-1668.
doi: 10.11999/JEIT171086
Abstract:
Under the premise of ensuring the security of Ciphertext-Policy Attribute Based Encryption (CP-ABE), to enhance efficiency as much as possible is always a research hotspot in the field of cryptography. Starting from the access structure, which is the efficiency basis of CP-ABE, a new kind of access structure is proposed based on Reduced Ordered Binary Decision Diagrams (ROBDD) for the first time, and the corresponding strategy representation method and satisfaction determination are given. Furthermore, based on the above access structure, a new CP-ABE with good performance in lots of aspects, such as time complexity of algorithms and storage occupancy of secret keys, is designed; In terms of security, the scheme can resist collusion attack and chosen plaintext attack. Comparative analysis shows that, ROBDD access structure has stronger expression ability and higher expression efficiency; In the new CP-ABE scheme, the time complexity of key generation algorithm and decryption algorithm is O(1), which can generate constant-size secret keys and achieve fast decryption.
Under the premise of ensuring the security of Ciphertext-Policy Attribute Based Encryption (CP-ABE), to enhance efficiency as much as possible is always a research hotspot in the field of cryptography. Starting from the access structure, which is the efficiency basis of CP-ABE, a new kind of access structure is proposed based on Reduced Ordered Binary Decision Diagrams (ROBDD) for the first time, and the corresponding strategy representation method and satisfaction determination are given. Furthermore, based on the above access structure, a new CP-ABE with good performance in lots of aspects, such as time complexity of algorithms and storage occupancy of secret keys, is designed; In terms of security, the scheme can resist collusion attack and chosen plaintext attack. Comparative analysis shows that, ROBDD access structure has stronger expression ability and higher expression efficiency; In the new CP-ABE scheme, the time complexity of key generation algorithm and decryption algorithm is O(1), which can generate constant-size secret keys and achieve fast decryption.
2018, 40(7): 1669-1675.
doi: 10.11999/JEIT170997
Abstract:
Why can fully homomorphic encryption be constructed based on lattice What is the essence and construction of the matrix An important concept is proposed: Abstract decryption structure. Based on the abstract decryption structure, the main factors related to the homomorphic encryption are analyzed and relationship between abstract decryption structure, homomorphism and noise control is studied. The construction of the homomorphic encryption is attributed to the problem of how to obtain the final decryption structure. So the formal method of homomorphic encryption can be established. Thus the essential law of the construction method of the homomorphic encryption construction is expounded, which provides the clue and clue for the construction of the new full homomorphic encryption. The general reason of the full homomorphic encryption of the ciphertext matrix from the point of view of the ciphertexts stack method is studied. The relation between the full homomorphic encryption and the other homomorphic encryption is obtained. Finally, this paper gives a general method of constructing fully homomorphic encryption.
Why can fully homomorphic encryption be constructed based on lattice What is the essence and construction of the matrix An important concept is proposed: Abstract decryption structure. Based on the abstract decryption structure, the main factors related to the homomorphic encryption are analyzed and relationship between abstract decryption structure, homomorphism and noise control is studied. The construction of the homomorphic encryption is attributed to the problem of how to obtain the final decryption structure. So the formal method of homomorphic encryption can be established. Thus the essential law of the construction method of the homomorphic encryption construction is expounded, which provides the clue and clue for the construction of the new full homomorphic encryption. The general reason of the full homomorphic encryption of the ciphertext matrix from the point of view of the ciphertexts stack method is studied. The relation between the full homomorphic encryption and the other homomorphic encryption is obtained. Finally, this paper gives a general method of constructing fully homomorphic encryption.
2018, 40(7): 1676-1683.
doi: 10.11999/JEIT170962
Abstract:
A malicious attack-resistant secure localization algorithm Evolutionary Location Algorithm with the Maximum Probability value (ELAMP) based on evolutionism is proposed. According to the maximum likelihood estimation probability model and the distribution of Received Signal Strength (RSS) standard deviation and the distance, a secure location model of ZigBee network is established. Furthermore, the evolutionary algorithm is designed to solve the model, and the convergence and the time complexity of the algorithm is analyzed. Experimental results show that the proposed algorithm has better positioning accuracy than the existing positioning algorithm when the proportion of malicious nodes is not more than 50%.
A malicious attack-resistant secure localization algorithm Evolutionary Location Algorithm with the Maximum Probability value (ELAMP) based on evolutionism is proposed. According to the maximum likelihood estimation probability model and the distribution of Received Signal Strength (RSS) standard deviation and the distance, a secure location model of ZigBee network is established. Furthermore, the evolutionary algorithm is designed to solve the model, and the convergence and the time complexity of the algorithm is analyzed. Experimental results show that the proposed algorithm has better positioning accuracy than the existing positioning algorithm when the proportion of malicious nodes is not more than 50%.
2018, 40(7): 1684-1690.
doi: 10.11999/JEIT170981
Abstract:
To improve the low acceptance ratio and revenue-cost ratio caused by the negligence of the topology attribute of the nodes in the existing virtual network embedding algorithm, the theory of fields in physics is introduced into the virtual network embedding, and a Virtual Network Embedding algorithm based on Topology Potential (TP-VNE) is proposed. In the node embedding stage, the virtual node is embedded onto the optimal physical node by calculating the topology potential of the node, the resource capacity of the node, and the distance between the embedded nodes and the node to embed. In the link embedding stage, the virtual link is embedded onto the best physical path by calculating the available bandwidth of the path and the hops of the path. Experimental results show that the proposed algorithm has the higher acceptance ratio and revenue-cost ratio compared with the existing virtual network embedding algorithm in all simulation conditions.
To improve the low acceptance ratio and revenue-cost ratio caused by the negligence of the topology attribute of the nodes in the existing virtual network embedding algorithm, the theory of fields in physics is introduced into the virtual network embedding, and a Virtual Network Embedding algorithm based on Topology Potential (TP-VNE) is proposed. In the node embedding stage, the virtual node is embedded onto the optimal physical node by calculating the topology potential of the node, the resource capacity of the node, and the distance between the embedded nodes and the node to embed. In the link embedding stage, the virtual link is embedded onto the best physical path by calculating the available bandwidth of the path and the hops of the path. Experimental results show that the proposed algorithm has the higher acceptance ratio and revenue-cost ratio compared with the existing virtual network embedding algorithm in all simulation conditions.
2018, 40(7): 1691-1698.
doi: 10.11999/JEIT170976
Abstract:
A delay minimization retransmission scheme based on an instantly decodable network coding is proposed to solve the conflict problem when multiple devices cooperatively retransmit in Device-to-Device (D2D) wireless networks concurrently. In retransmission stage, making full use of multiple devices cooperative transmission advantages in D2D wireless network, combined with the packet receiving state of each devices, taking all of the influence factors of delay into account, and then the packets with smaller incremental delay for each retransmission are selected to generate encoding packets to minimize the retransmission delay. At the same time, the devices conflict graph is constructed and the maximal independent set is searched in the graph. According to the encoding package weight value of each device, the maximum weighted independent set are selected as the concurrent cooperative retransmission devices to reduce the number of retransmission. Simulation results show that the proposed scheme can further improve the retransmission efficiency of D2D wireless network.
A delay minimization retransmission scheme based on an instantly decodable network coding is proposed to solve the conflict problem when multiple devices cooperatively retransmit in Device-to-Device (D2D) wireless networks concurrently. In retransmission stage, making full use of multiple devices cooperative transmission advantages in D2D wireless network, combined with the packet receiving state of each devices, taking all of the influence factors of delay into account, and then the packets with smaller incremental delay for each retransmission are selected to generate encoding packets to minimize the retransmission delay. At the same time, the devices conflict graph is constructed and the maximal independent set is searched in the graph. According to the encoding package weight value of each device, the maximum weighted independent set are selected as the concurrent cooperative retransmission devices to reduce the number of retransmission. Simulation results show that the proposed scheme can further improve the retransmission efficiency of D2D wireless network.
2018, 40(7): 1699-1706.
doi: 10.11999/JEIT170930
Abstract:
In order to improve the overall effectiveness of the online assignment of crowdsourcing tasks, an online task assignment method is proposed for the space-time crowdsourcing environment. To deal with the problem of online task assignment in spatiotemporal crowdsourcing environment, a K-NearestNeighbor (KNN) algorithm is firstly proposed based on crowdsourcing task to select the candidate crowdsourcing workers. Then a threshold selection algorithm based on dynamic utility is designed to realize the optimal allocation of crowdsourcing workers and tasks. Experimental results show that the proposed algorithm is effective and feasible, and can guarantee the reliability of crowdsourcing workers and optimize the overall efficiency of the platform.
In order to improve the overall effectiveness of the online assignment of crowdsourcing tasks, an online task assignment method is proposed for the space-time crowdsourcing environment. To deal with the problem of online task assignment in spatiotemporal crowdsourcing environment, a K-NearestNeighbor (KNN) algorithm is firstly proposed based on crowdsourcing task to select the candidate crowdsourcing workers. Then a threshold selection algorithm based on dynamic utility is designed to realize the optimal allocation of crowdsourcing workers and tasks. Experimental results show that the proposed algorithm is effective and feasible, and can guarantee the reliability of crowdsourcing workers and optimize the overall efficiency of the platform.
2018, 40(7): 1707-1714.
doi: 10.11999/JEIT170966
Abstract:
In order to control the information propagation of the whole network at a lower cost, some information propagation control methods are introduced into social networks to select the best control point at a proper time. However, few work considers the weak ties between nodes to control the information propagation. Due to the characteristics of the complementation of information demand and the continuous assimilation of behavior orientation, the weak ties between nodes may be explosive in the process of information propagation, thus they can not be ignored. To solve this problem, considering the impact of strong and weak ties between nodes on information propagation, a propagation control method based on the exact controllability theory is proposed. Firstly, some strong ties between nodes, such as the node's intimacy, authority and interaction frequency are introduced to build the initial tie networks. Secondly, some potential valuable weak ties between nodes are identified and then tie networks are further updated. Finally, the exact controllability theory is used to find the driver node groups, and then the set of driver nodes are selected according to the characteristics of information propagation to control information propagation. Experimental results show that the proposed method can effectively promote or suppress the information propagation, which provides some ideas for the information propagation control in social networks.
In order to control the information propagation of the whole network at a lower cost, some information propagation control methods are introduced into social networks to select the best control point at a proper time. However, few work considers the weak ties between nodes to control the information propagation. Due to the characteristics of the complementation of information demand and the continuous assimilation of behavior orientation, the weak ties between nodes may be explosive in the process of information propagation, thus they can not be ignored. To solve this problem, considering the impact of strong and weak ties between nodes on information propagation, a propagation control method based on the exact controllability theory is proposed. Firstly, some strong ties between nodes, such as the node's intimacy, authority and interaction frequency are introduced to build the initial tie networks. Secondly, some potential valuable weak ties between nodes are identified and then tie networks are further updated. Finally, the exact controllability theory is used to find the driver node groups, and then the set of driver nodes are selected according to the characteristics of information propagation to control information propagation. Experimental results show that the proposed method can effectively promote or suppress the information propagation, which provides some ideas for the information propagation control in social networks.
2018, 40(7): 1715-1722.
doi: 10.11999/JEIT170927
Abstract:
To solve the problem that the network capacity decreases with the increasing interference in Wireless Sensor Networks (WSNs), a joint power control and channel allocation optimization game model is constructed, which considers the limitation of network energy. This game model contains the network capacity and the energy consumption of data transmission in the network. Theoretical analysis proves the existence of the optimal power and the optimal channel. Based on the model, a joint Power control and Channel allocation Optimization Algorithm for wireless sensor networks (PCOA) is proposed, which adopts the best response strategy. The theoretical analysis proves that this algorithm can converge to Nash Equilibrium. Besides, the information complexity of this algorithm is small. Simulation results show that PCOA algorithm can reduce the interference and the energy consumption, which increases the network capacity.
To solve the problem that the network capacity decreases with the increasing interference in Wireless Sensor Networks (WSNs), a joint power control and channel allocation optimization game model is constructed, which considers the limitation of network energy. This game model contains the network capacity and the energy consumption of data transmission in the network. Theoretical analysis proves the existence of the optimal power and the optimal channel. Based on the model, a joint Power control and Channel allocation Optimization Algorithm for wireless sensor networks (PCOA) is proposed, which adopts the best response strategy. The theoretical analysis proves that this algorithm can converge to Nash Equilibrium. Besides, the information complexity of this algorithm is small. Simulation results show that PCOA algorithm can reduce the interference and the energy consumption, which increases the network capacity.
2018, 40(7): 1723-1730.
doi: 10.11999/JEIT170972
Abstract:
The k-error linear complexity of a sequence is a fundamental concept for assessing the stability of the linear complexity. After computing the k-error linear complexity of a sequence, those bits that make the linear complexity reduced also need to be computed. For 2pn-periodic sequence over GF(q) , where p and q are odd primes and q is a primitive root modulo p2, an algorithm is presented, which not only computes the k-error linear complexity of a sequence s but also gets the corresponding error sequence e. A function is designed to trace the vector cost called “trace function”, so the error sequence e can be computed by calling the “trace function”, and the linear complexity of (s+e) reaches the k-error linear complexity of the sequence s.
The k-error linear complexity of a sequence is a fundamental concept for assessing the stability of the linear complexity. After computing the k-error linear complexity of a sequence, those bits that make the linear complexity reduced also need to be computed. For 2pn-periodic sequence over GF(q) , where p and q are odd primes and q is a primitive root modulo p2, an algorithm is presented, which not only computes the k-error linear complexity of a sequence s but also gets the corresponding error sequence e. A function is designed to trace the vector cost called “trace function”, so the error sequence e can be computed by calling the “trace function”, and the linear complexity of (s+e) reaches the k-error linear complexity of the sequence s.
2018, 40(7): 1731-1737.
doi: 10.11999/JEIT170895
Abstract:
In order to solve the limitation of processing capacity and energy of single mobile equipment, the conception of Ad-hoc mobile cloud is proposed recently, in which a mobile device can use the idle resources at other neighboring devices for processing data and storage in Ad-hoc manner. To this end, this paper designs a workload distribution for offloading among mobile equipment. Considering the random and intermittent connections between mobile equipment caused by the movement in wireless network, a stochastic programming method is adopted to take posterior recourse actions to compensate for inaccurate predictions. Moreover, in order to motivate the available mobile equipment for offloading while maximizing their utilities, a distributed multi-stage Stochastic buyer/seller Game for Workload Distribution (SGWD) is formulated. Numerical results show the effectiveness of SGWD compared with the benchmark method in terms of communication cost, the delay, energy consumption and the payoff.
2018, 40(7): 1738-1744.
doi: 10.11999/JEIT170902
Abstract:
In order to solve the problem of resource allocation between 5G virtual network slice, a resource scheduling mechanism based on Online Double Auction (ODA) is proposed. Firstly, the priority of network slices and unit resource quotes are determined according to different traffic needs and traffic benefits. Then, to maximize the network revenue, an offline single auction model is established. Further, based on the resources dynamic allocation and recycling, the price-updating algorithm is proposed to update the resource price in real time. Finally, the offline single auction mechanism and the price update mechanism are combined to establish ODA model and allocate resources dynamically for the network slices. The simulation results show that the proposed mechanism can improve network revenue and guarantee the QoS requirement of each slice user.
In order to solve the problem of resource allocation between 5G virtual network slice, a resource scheduling mechanism based on Online Double Auction (ODA) is proposed. Firstly, the priority of network slices and unit resource quotes are determined according to different traffic needs and traffic benefits. Then, to maximize the network revenue, an offline single auction model is established. Further, based on the resources dynamic allocation and recycling, the price-updating algorithm is proposed to update the resource price in real time. Finally, the offline single auction mechanism and the price update mechanism are combined to establish ODA model and allocate resources dynamically for the network slices. The simulation results show that the proposed mechanism can improve network revenue and guarantee the QoS requirement of each slice user.
2018, 40(7): 1745-1751.
doi: 10.11999/JEIT170952
Abstract:
Tropospheric scatter (Troposcatter) communication is an important means for ground microwave beyond-line-of-sight propagation. Available troposcatter transmission loss models are inefficient to describe the random variables resulting from atmosphere environment and other factors. Therefore, this paper studies the short-term fading and long-term fading characteristics of transmission loss based on those of electric field strength for the first time. The distribution model of transmission loss is modeled, whose parameters are estimated referring to ITU-R P.617-3. Parts of measured scatter links data from International Telecommunication Union (ITU) are chosen to verify this model with normal distribution graph. The result shows that the long-term fading of transmission obeys the normal distribution. It gives the basis of calculating error bit for the further work. In addition, a transmission loss prediction method is proposed based on its distribution model. It is verified to have a good accuracy using measured data, and this method address the problem that available transmission loss methods can not predict the values at any time percentages.
Tropospheric scatter (Troposcatter) communication is an important means for ground microwave beyond-line-of-sight propagation. Available troposcatter transmission loss models are inefficient to describe the random variables resulting from atmosphere environment and other factors. Therefore, this paper studies the short-term fading and long-term fading characteristics of transmission loss based on those of electric field strength for the first time. The distribution model of transmission loss is modeled, whose parameters are estimated referring to ITU-R P.617-3. Parts of measured scatter links data from International Telecommunication Union (ITU) are chosen to verify this model with normal distribution graph. The result shows that the long-term fading of transmission obeys the normal distribution. It gives the basis of calculating error bit for the further work. In addition, a transmission loss prediction method is proposed based on its distribution model. It is verified to have a good accuracy using measured data, and this method address the problem that available transmission loss methods can not predict the values at any time percentages.
2018, 40(7): 1752-1758.
doi: 10.11999/JEIT170844
Abstract:
A construction of Gaussian integer sequences based on pseudo-random sequences. Gaussian integer sequences with period pm-1 whose degree p-1 are constructed from p-ary pseudo-random sequences with period pm-1. The presented sequences are nearly perfect Gaussian integer sequences with p-2 non-zero out-of-phase autocorrelation values. Moreover, these Gaussian integer sequences have balance property, as a result, they will be widely used in wireless communication and radar systems.
A construction of Gaussian integer sequences based on pseudo-random sequences. Gaussian integer sequences with period pm-1 whose degree p-1 are constructed from p-ary pseudo-random sequences with period pm-1. The presented sequences are nearly perfect Gaussian integer sequences with p-2 non-zero out-of-phase autocorrelation values. Moreover, these Gaussian integer sequences have balance property, as a result, they will be widely used in wireless communication and radar systems.
2018, 40(7): 1759-1766.
doi: 10.11999/JEIT170989
Abstract:
Near-threshold voltage computing enables transistor voltage scaling to continue with Moore’s Law projection and dramatically improves power and energy efficiency. However, a great number of bit-cell errors occur in large SRAM structures, such as Last-Level Cache (LLC). A Fault-Tolerant LLC (FTLLC) design with conventional 6T SRAM cells is proposed to deal with a higher failure rate which is more than 1% at near-threshold voltage. FTLLC improves the reliability of data stored in Cache by correcting the single-error and compressing multi-errors in Cache entry. To validate the efficiency of FTLLC, FTLLC and prior works are implemented in gem5, and are simulated with SPEC CPU2006. The experiment shows that compared with Concertina at 650 mV, the performance of a 65 nm FTLLC with 4-Byte subblock size improves by 7.2% and the Cache capacity increases by 24.9%. Besides, the miss rate decreases by 58.2%, and there are little increases on area overhead and power consumption.
Near-threshold voltage computing enables transistor voltage scaling to continue with Moore’s Law projection and dramatically improves power and energy efficiency. However, a great number of bit-cell errors occur in large SRAM structures, such as Last-Level Cache (LLC). A Fault-Tolerant LLC (FTLLC) design with conventional 6T SRAM cells is proposed to deal with a higher failure rate which is more than 1% at near-threshold voltage. FTLLC improves the reliability of data stored in Cache by correcting the single-error and compressing multi-errors in Cache entry. To validate the efficiency of FTLLC, FTLLC and prior works are implemented in gem5, and are simulated with SPEC CPU2006. The experiment shows that compared with Concertina at 650 mV, the performance of a 65 nm FTLLC with 4-Byte subblock size improves by 7.2% and the Cache capacity increases by 24.9%. Besides, the miss rate decreases by 58.2%, and there are little increases on area overhead and power consumption.
2018, 40(7): 1767-1773.
doi: 10.11999/JEIT170998
Abstract:
The design and experimental verification of the launcher and mirror system of W band TE62 mode gyrotron quasi optical mode converter are presented. Based on the coupled mode theory, two order perturbation is used to design the launcher. The field distribution of the Gauss beam spot on the inner wall of the circular waveguide is obtained. The vector diffraction integral theory based on Huygens,s principle is used to optimize the mirror system of the optical mode converter. Simulation and calculation results show that the mode conversion efficiency of quasi optical mode the converter is 92.3%. Finally, a thermal measurement experiment is carried out to verify that the output mode is W band Gauss like mode.
The design and experimental verification of the launcher and mirror system of W band TE62 mode gyrotron quasi optical mode converter are presented. Based on the coupled mode theory, two order perturbation is used to design the launcher. The field distribution of the Gauss beam spot on the inner wall of the circular waveguide is obtained. The vector diffraction integral theory based on Huygens,s principle is used to optimize the mirror system of the optical mode converter. Simulation and calculation results show that the mode conversion efficiency of quasi optical mode the converter is 92.3%. Finally, a thermal measurement experiment is carried out to verify that the output mode is W band Gauss like mode.
2018, 40(7): 1774-1778.
doi: 10.11999/JEIT170834
Abstract:
The most important characteristic of the complex electromagnetic environment is the limitation of the radio frequency resource. As a result, the smart use of the limiting spectral resource is necessary to the waveform design for the cognitive radar. This paper designs the transmit waveform under the Peak to Average power Ratio (PAR), to maximize the Signal to Noise power Ratio (SNR) at the receiver, and simultaneously, minimize the power of the waveform in the interference frequency bands. The waveform design problem is a quadratically constrained multi-objective optimization problem. Exploiting the Pareto optimization method, the one objective function is obtained by weighted sum of the two ones, and the resultant problem reduces into a Quadratically Constrained Quadratic Program (QCQP). In order to solve it, the SemiDefinite Program (SDP) relaxation and randomization are used to achieve the optimal waveform, whose performance is related to the Pareto weights and the PAR constraint. The computer simulation results show that, there is a restrictive relationship between the SNR and interference suppression ability for the waveform design, and the performance can be improved by increasing the dynamic range of the transmitter.
The most important characteristic of the complex electromagnetic environment is the limitation of the radio frequency resource. As a result, the smart use of the limiting spectral resource is necessary to the waveform design for the cognitive radar. This paper designs the transmit waveform under the Peak to Average power Ratio (PAR), to maximize the Signal to Noise power Ratio (SNR) at the receiver, and simultaneously, minimize the power of the waveform in the interference frequency bands. The waveform design problem is a quadratically constrained multi-objective optimization problem. Exploiting the Pareto optimization method, the one objective function is obtained by weighted sum of the two ones, and the resultant problem reduces into a Quadratically Constrained Quadratic Program (QCQP). In order to solve it, the SemiDefinite Program (SDP) relaxation and randomization are used to achieve the optimal waveform, whose performance is related to the Pareto weights and the PAR constraint. The computer simulation results show that, there is a restrictive relationship between the SNR and interference suppression ability for the waveform design, and the performance can be improved by increasing the dynamic range of the transmitter.
