Email alert
2022 Vol. 44, No. 12
Display Method:
2022, 44(12): 4075-4084.
doi: 10.11999/JEIT220363
Abstract:
Computational microwave imaging based on metamaterial aperture can be considered as microwave compression sensing imaging. The imaging effect of this imaging method is seriously affected by the grid mismatch error. In this paper, a Two-Dimensional (2D) off-grid observation model based on Sinc interpolation function is constructed by analyzing the reconstruction process of 2D scene in the computational microwave imaging system for metamaterial aperture. On this basis, an Off-Grid imaging method using Sinc Interpolation based on Sparse Bayesian Learning (OGSISBL) is proposed. Under the framework of the expectation maximization algorithm, the amplitude and position of the return of the scatterers are recovered, and the off-grid error is calibrated. The performance of the proposed algorithm is verified by imaging the simulation data of the computing microwave imaging system based on metamaterial aperture. The results show that the proposed algorithm has strong robustness.
Computational microwave imaging based on metamaterial aperture can be considered as microwave compression sensing imaging. The imaging effect of this imaging method is seriously affected by the grid mismatch error. In this paper, a Two-Dimensional (2D) off-grid observation model based on Sinc interpolation function is constructed by analyzing the reconstruction process of 2D scene in the computational microwave imaging system for metamaterial aperture. On this basis, an Off-Grid imaging method using Sinc Interpolation based on Sparse Bayesian Learning (OGSISBL) is proposed. Under the framework of the expectation maximization algorithm, the amplitude and position of the return of the scatterers are recovered, and the off-grid error is calibrated. The performance of the proposed algorithm is verified by imaging the simulation data of the computing microwave imaging system based on metamaterial aperture. The results show that the proposed algorithm has strong robustness.
2022, 44(12): 4085-4094.
doi: 10.11999/JEIT220539
Abstract:
This paper presents an electromagnetic metasurface structure with partially polarization-conversion and partially reflection functions and its application to the Fabry-Perot (F-P) cavity antenna design with wideband, high-gain and Circularly Polarized (CP) features. On one hand, when the metasurface is backed by a reflection ground, it is able to show the partially polarization-conversion function and can be applied to the design of the CP radiation source of the F-P antenna. On the other hand, when the reflection ground is removed, the metasurface exhibits partially reflection performance, and thus it can be used as the Partially Reflection Surface (PRS). By placing a rectangle patch upon the Partially Polarization-Conversion Surface (PPCS) and loading parasitic patches and the PRS, the linearly-polarized radiation from the source patch can be transformed into the CP radiation. In addition, the impedance and Axial Ratio (AR) bandwidths of the antenna are both enhanced. The designed antenna is simulated, fabricated and measured. Measured results indicate that the impedance and AR bandwidths are 6.8 ～ 8.4 GHz (21.3%) and 6.8 ～ 8.3 GHz (19.9%), respectively. Besides, the peak realized gain reaches 10.5 dBi.
This paper presents an electromagnetic metasurface structure with partially polarization-conversion and partially reflection functions and its application to the Fabry-Perot (F-P) cavity antenna design with wideband, high-gain and Circularly Polarized (CP) features. On one hand, when the metasurface is backed by a reflection ground, it is able to show the partially polarization-conversion function and can be applied to the design of the CP radiation source of the F-P antenna. On the other hand, when the reflection ground is removed, the metasurface exhibits partially reflection performance, and thus it can be used as the Partially Reflection Surface (PRS). By placing a rectangle patch upon the Partially Polarization-Conversion Surface (PPCS) and loading parasitic patches and the PRS, the linearly-polarized radiation from the source patch can be transformed into the CP radiation. In addition, the impedance and Axial Ratio (AR) bandwidths of the antenna are both enhanced. The designed antenna is simulated, fabricated and measured. Measured results indicate that the impedance and AR bandwidths are 6.8 ～ 8.4 GHz (21.3%) and 6.8 ～ 8.3 GHz (19.9%), respectively. Besides, the peak realized gain reaches 10.5 dBi.
2022, 44(12): 4095-4103.
doi: 10.11999/JEIT220007
Abstract:
This paper proposes a novel design strategy for wideband ultralow-profile Folded TransmitArray (FTA) antenna using metasurfaces. It consists of two kinds of metasurfaces and an open waveguide antenna as the feed source. The bottom metasurface is applied to transforming the linearly polarized wave coming from feed antenna into its cross-polarized reflected ones. Meanwhile, the top metasurface is capable of reflecting one linearly polarized wave and transmitting the other orthogonal counterpart. With smart design, the proposed FTA can reflect the electromagnetic wave back three times between two metasurfaces and realize high gain performance within a wide frequency band, and its profile height is decreased to 1/4 that of conventional transmitarray. The simulated and measured results are basically consistent, which demonstrates a 3 dB gain bandwidth of 19.6% (9.2～11.2 GHz), a peak gain of 21 dBi and a peak aperture efficiency of 30% at 9.6 GHz. The proposed design offers a new avenue for wideband low profile array antenna.
This paper proposes a novel design strategy for wideband ultralow-profile Folded TransmitArray (FTA) antenna using metasurfaces. It consists of two kinds of metasurfaces and an open waveguide antenna as the feed source. The bottom metasurface is applied to transforming the linearly polarized wave coming from feed antenna into its cross-polarized reflected ones. Meanwhile, the top metasurface is capable of reflecting one linearly polarized wave and transmitting the other orthogonal counterpart. With smart design, the proposed FTA can reflect the electromagnetic wave back three times between two metasurfaces and realize high gain performance within a wide frequency band, and its profile height is decreased to 1/4 that of conventional transmitarray. The simulated and measured results are basically consistent, which demonstrates a 3 dB gain bandwidth of 19.6% (9.2～11.2 GHz), a peak gain of 21 dBi and a peak aperture efficiency of 30% at 9.6 GHz. The proposed design offers a new avenue for wideband low profile array antenna.
2022, 44(12): 4104-4110.
doi: 10.11999/JEIT211057
Abstract:
In this paper, to simplify the control circuit of the Reconfigurable TransmitArray(RTA), a RTA based on row-column beamsteering method is proposed, which is composed of a double-layer Frequency Selective Surfaces(FSS). By tunning a capacitance value of the varactor intergrated in the frequency selective surface element, the RTA element can tune the transmission phase. At the same time, a row-column beamsteering method is used to tune a Direction Current (DC) bias voltage between the ports of the varactor, while a line can tune the DC bias voltage at each row elements or each column elements of the RTA. Due to the limited phase range of the RTA element, a phase-correction method is also utilized to reduce the elements phase error during 2-D beamscanning. The scanned beam results show that the scanned beam angle can reach 39° with –1.7 dB gain loss in E-plane and can reach 33° with –3 dB gain loss in H-plane. The RTA has the advantages of simple 2-D beam-steering circuit and low cost, the proposed RTA has great potential for radar system and modern communication system application.
In this paper, to simplify the control circuit of the Reconfigurable TransmitArray(RTA), a RTA based on row-column beamsteering method is proposed, which is composed of a double-layer Frequency Selective Surfaces(FSS). By tunning a capacitance value of the varactor intergrated in the frequency selective surface element, the RTA element can tune the transmission phase. At the same time, a row-column beamsteering method is used to tune a Direction Current (DC) bias voltage between the ports of the varactor, while a line can tune the DC bias voltage at each row elements or each column elements of the RTA. Due to the limited phase range of the RTA element, a phase-correction method is also utilized to reduce the elements phase error during 2-D beamscanning. The scanned beam results show that the scanned beam angle can reach 39° with –1.7 dB gain loss in E-plane and can reach 33° with –3 dB gain loss in H-plane. The RTA has the advantages of simple 2-D beam-steering circuit and low cost, the proposed RTA has great potential for radar system and modern communication system application.
2022, 44(12): 4111-4115.
doi: 10.11999/JEIT220569
Abstract:
In this paper, a new type of foam material PolyMethacrylimIde (PMI) is used to design a millimeter-wave Luneburg lens antenna for the detection of complex space environment. By correlating the density of the foam material with the dielectric constant, combined with the working principle of the traditional Luneburg lens antenna, the simulation optimization is carried out, and the function of miniaturized high-gain multi-beam is realized. The simulation results show that the antenna works at 33.7 GHz, the gain can reach 25.65 dBi, and the beam width is 4.17°. This design method provides a new idea for the realization of miniaturized high-gain Luneburg lenses in the future.
In this paper, a new type of foam material PolyMethacrylimIde (PMI) is used to design a millimeter-wave Luneburg lens antenna for the detection of complex space environment. By correlating the density of the foam material with the dielectric constant, combined with the working principle of the traditional Luneburg lens antenna, the simulation optimization is carried out, and the function of miniaturized high-gain multi-beam is realized. The simulation results show that the antenna works at 33.7 GHz, the gain can reach 25.65 dBi, and the beam width is 4.17°. This design method provides a new idea for the realization of miniaturized high-gain Luneburg lenses in the future.
2022, 44(12): 4116-4124.
doi: 10.11999/JEIT220447
Abstract:
Polarization conversion has important research significance and application value in the field of terahertz modulation. Traditional polarization conversion devices have many shortcomings, such as large size, low integration, high loss and narrow bandwidth. In this paper, a symmetrical "mountain" structure of resonator is proposed, which can be used to realize the design of reflection and transmission polarization conversion devices. The reflective device realizes the linear polarization conversion with the properties of broadband and extremely high polarization conversion rate. The transmissive device realizes 135.5% ultra-broadband linear polarization conversion. The anisotropy theory is used to analyze the mechanism of polarization conversion in reflective devices, and the Fabry-Pérot-like cavity formed by the resonant structure array and the metallized ground plane is calculated based on the multiple interference theory. The calculated results are in good agreement with the simulations. Furthermore, the Fabry-Pérot-like cavity composed of orthogonal grating and the resonator are used to form a transmissive device. The contributions of different parts of the resonator to the broadband polarization conversion are analyzed. Contribution of different structures validates results for broadband polarization conversion. The research results provide a new idea for the realization of ultra-wideband polarization conversion devices based on fixed phase difference and the application of Fabry-Pérot-like cavities in metasurfaces.
Polarization conversion has important research significance and application value in the field of terahertz modulation. Traditional polarization conversion devices have many shortcomings, such as large size, low integration, high loss and narrow bandwidth. In this paper, a symmetrical "mountain" structure of resonator is proposed, which can be used to realize the design of reflection and transmission polarization conversion devices. The reflective device realizes the linear polarization conversion with the properties of broadband and extremely high polarization conversion rate. The transmissive device realizes 135.5% ultra-broadband linear polarization conversion. The anisotropy theory is used to analyze the mechanism of polarization conversion in reflective devices, and the Fabry-Pérot-like cavity formed by the resonant structure array and the metallized ground plane is calculated based on the multiple interference theory. The calculated results are in good agreement with the simulations. Furthermore, the Fabry-Pérot-like cavity composed of orthogonal grating and the resonator are used to form a transmissive device. The contributions of different parts of the resonator to the broadband polarization conversion are analyzed. Contribution of different structures validates results for broadband polarization conversion. The research results provide a new idea for the realization of ultra-wideband polarization conversion devices based on fixed phase difference and the application of Fabry-Pérot-like cavities in metasurfaces.
2022, 44(12): 4125-4133.
doi: 10.11999/JEIT211111
Abstract:
In order to meet the demand for passive ranging of sound sources in shallow water, a normal mode separation and ranging method based on a large-aperture horizontal array is proposed. This method deals with the problem of modal separation that caused by the bending of the modal curve in the frequency-wavenumber domain. Under the condition that the cut-off frequency of each order normal mode does not change with the signal frequency, a alignment method based on wavenumber scaling is presented to realize the effective separation of normal modes. The energy focusing of normal modes is realized by nonlinear phase compensation, and the passive ranging of the sound source is realized by combining distance traversal and peak extraction. This method can effectively achieve gains in space and frequency domains, and obtain multi-modal energy accumulation, which provides a new approach for the distance estimation of weak sound sources. The effectiveness of the method is verified by simulation data.
In order to meet the demand for passive ranging of sound sources in shallow water, a normal mode separation and ranging method based on a large-aperture horizontal array is proposed. This method deals with the problem of modal separation that caused by the bending of the modal curve in the frequency-wavenumber domain. Under the condition that the cut-off frequency of each order normal mode does not change with the signal frequency, a alignment method based on wavenumber scaling is presented to realize the effective separation of normal modes. The energy focusing of normal modes is realized by nonlinear phase compensation, and the passive ranging of the sound source is realized by combining distance traversal and peak extraction. This method can effectively achieve gains in space and frequency domains, and obtain multi-modal energy accumulation, which provides a new approach for the distance estimation of weak sound sources. The effectiveness of the method is verified by simulation data.
2022, 44(12): 4134-4143.
doi: 10.11999/JEIT211009
Abstract:
The rapid development of Hypersonic Glide Vehicle (HGV) has changed the traditional combat style and opened a new field of military struggle. Identifying the maneuvering state of HGV can provide a powerful support for threat assessment, trajectory prediction and defense decision. In order to improve the accuracy of HGV maneuver state recognition, an HGV maneuver state recognition model based on ATtention Convolutional Long Short-Term Memory network (AT-ConvLSTM) is proposed. First, on the basis of maneuvering modeling and characteristic analysis of HGV, the maneuvering state of HGV in space is divided into eight categories, and the corresponding feature recognition parameters are constructed. A trajectory library containing HGV maneuvering trajectories under different initial conditions and control modes is established. Then, the conversion steps from radar tracking information to feature recognition parameters are deduced. The proposed state recognition model is used to extract the spatial features of HGV motion trajectory, and the maneuvering state is classified by the SoftMax classifier. Finally, the algorithm is verified by simulation experiments. The results show that the proposed method can effectively identify HGV maneuvering state online, which has good real-time and accuracy.
The rapid development of Hypersonic Glide Vehicle (HGV) has changed the traditional combat style and opened a new field of military struggle. Identifying the maneuvering state of HGV can provide a powerful support for threat assessment, trajectory prediction and defense decision. In order to improve the accuracy of HGV maneuver state recognition, an HGV maneuver state recognition model based on ATtention Convolutional Long Short-Term Memory network (AT-ConvLSTM) is proposed. First, on the basis of maneuvering modeling and characteristic analysis of HGV, the maneuvering state of HGV in space is divided into eight categories, and the corresponding feature recognition parameters are constructed. A trajectory library containing HGV maneuvering trajectories under different initial conditions and control modes is established. Then, the conversion steps from radar tracking information to feature recognition parameters are deduced. The proposed state recognition model is used to extract the spatial features of HGV motion trajectory, and the maneuvering state is classified by the SoftMax classifier. Finally, the algorithm is verified by simulation experiments. The results show that the proposed method can effectively identify HGV maneuvering state online, which has good real-time and accuracy.
2022, 44(12): 4144-4150.
doi: 10.11999/JEIT210942
Abstract:
There are some mismatches for the estimation of signal parameters in mult-dimention Estimating Signal Parameter via Rotational Invariance Techniques (ESPRIT) algorithm. In this paper, a parameter pairing method based on eigenvalue fractal dimension is proposed. By linearly combining the eigenvalues and constructing a judgment matrix, the parameters are paired according to the corresponding relationship of matrix dimension. In contrast to other pairing algorithms, the proposed algorithm can achieve automatic pairing with low complexity, non-fuzzy parameters and higher robustness. Several examples are given to demonstrate the effectiveness of the proposed algorithm.
There are some mismatches for the estimation of signal parameters in mult-dimention Estimating Signal Parameter via Rotational Invariance Techniques (ESPRIT) algorithm. In this paper, a parameter pairing method based on eigenvalue fractal dimension is proposed. By linearly combining the eigenvalues and constructing a judgment matrix, the parameters are paired according to the corresponding relationship of matrix dimension. In contrast to other pairing algorithms, the proposed algorithm can achieve automatic pairing with low complexity, non-fuzzy parameters and higher robustness. Several examples are given to demonstrate the effectiveness of the proposed algorithm.
2022, 44(12): 4151-4158.
doi: 10.11999/JEIT201061
Abstract:
For the problem of Phased Array Radar (PAR) has excessive resource consumption and low measurement accuracy in the process of detecting Hypersonic Gliding Vehicle (HGV). An adaptive radar power allocation method based Radar Cross-Section (RCS) prediction is presented in this paper. Based on the target state and RCS information in the sliding window, Bayesian posterior probability formula is utilized to predict the target RCS at the next moment. Then, the transmit pulse dwell time is adjusted to achieve dynamic adjustment of radar resources, so that the target echo signal signal-to-noise ratio remains stable and the radar tracking performance is improved. The simulation experiment shows that the method in this paper can make accurate prediction the RCS of next time then adaptive allocates radar power. To achieve the purpose of improving the tracking accuracy under the conditions of radar resources.
For the problem of Phased Array Radar (PAR) has excessive resource consumption and low measurement accuracy in the process of detecting Hypersonic Gliding Vehicle (HGV). An adaptive radar power allocation method based Radar Cross-Section (RCS) prediction is presented in this paper. Based on the target state and RCS information in the sliding window, Bayesian posterior probability formula is utilized to predict the target RCS at the next moment. Then, the transmit pulse dwell time is adjusted to achieve dynamic adjustment of radar resources, so that the target echo signal signal-to-noise ratio remains stable and the radar tracking performance is improved. The simulation experiment shows that the method in this paper can make accurate prediction the RCS of next time then adaptive allocates radar power. To achieve the purpose of improving the tracking accuracy under the conditions of radar resources.
2022, 44(12): 4159-4167.
doi: 10.11999/JEIT211120
Abstract:
Considering the problem of the low resolution of the Capon Power Spectrum (CPS) in the reconstruction of Interference plus Noise Covariance Matrix (INCM), two Robust Adaptive Beamforming (RAB) algorithms are proposed. The proposed algorithm first searches the peaks of CPS to determine the integration intervals and then eigen-decomposes the covariance matrixes obtained from the integration of each interval. The number of incident sources in the interval is determined by reasonably setting the decision threshold, and the eigenvectors corresponding to the larger eigenvalues are used as the preliminary estimation of the Steering Vectors (SV). Then, by maximizing the estimated power, the gap between the nominal SV and the real SV is searched in the orthogonal space of the nominal SV. The first proposed algorithm uses the eigenvector corresponding to the minimum eigenvalue to add the orthogonal proportional gradient to the initial estimated SV to obtain the double-layer estimated SV. The second proposed algorithm obtains the modified SV by solving a Quadratic Programming (QP) problem. Finally, the optimal weight vector of the array is obtained by reconstructing the INCM. Simulation results demonstrate that the proposed algorithm solves effectively the problem of the low resolution of the CPS estimation and is superior to other algorithms.
Considering the problem of the low resolution of the Capon Power Spectrum (CPS) in the reconstruction of Interference plus Noise Covariance Matrix (INCM), two Robust Adaptive Beamforming (RAB) algorithms are proposed. The proposed algorithm first searches the peaks of CPS to determine the integration intervals and then eigen-decomposes the covariance matrixes obtained from the integration of each interval. The number of incident sources in the interval is determined by reasonably setting the decision threshold, and the eigenvectors corresponding to the larger eigenvalues are used as the preliminary estimation of the Steering Vectors (SV). Then, by maximizing the estimated power, the gap between the nominal SV and the real SV is searched in the orthogonal space of the nominal SV. The first proposed algorithm uses the eigenvector corresponding to the minimum eigenvalue to add the orthogonal proportional gradient to the initial estimated SV to obtain the double-layer estimated SV. The second proposed algorithm obtains the modified SV by solving a Quadratic Programming (QP) problem. Finally, the optimal weight vector of the array is obtained by reconstructing the INCM. Simulation results demonstrate that the proposed algorithm solves effectively the problem of the low resolution of the CPS estimation and is superior to other algorithms.
2022, 44(12): 4168-4177.
doi: 10.11999/JEIT211138
Abstract:
For the problem that the results of multi-sensor measurement iteratively updating Gaussian Mixture Probability Hypothesis Density (GM-PHD) tracking algorithm is sensitive to the updating order if the qualities of multi-sensor observation data are different and unknown, a multi-sensor Adaptive observation Iteratively Updating GM-PHD tracking algorithm (AIU-GM-PHD) is proposed. Firstly, based on the multi-sensor fusion consistency measure, a method is proposed to evaluate the online quality of each sensor's tracking results. Then, the sequence of multi-sensor iterative fusion is optimized. Finally, the corresponding multi-sensor GM-PHD fusion tracking algorithm is constructed. To solve the problem that the multi-sensor adaptive order iterative fusion can not reflect the sensor quality gap, an Adaptive Iteratively Updating GM-PHD tracking algorithm PAIU-GM-PHD with weighted pseudo measurements is proposed. The simulation results show that, compared with the conventional multi-sensor iterative update GM-PHD tracking algorithm, the proposed algorithms can obtain more robust and accurate tracking results.
For the problem that the results of multi-sensor measurement iteratively updating Gaussian Mixture Probability Hypothesis Density (GM-PHD) tracking algorithm is sensitive to the updating order if the qualities of multi-sensor observation data are different and unknown, a multi-sensor Adaptive observation Iteratively Updating GM-PHD tracking algorithm (AIU-GM-PHD) is proposed. Firstly, based on the multi-sensor fusion consistency measure, a method is proposed to evaluate the online quality of each sensor's tracking results. Then, the sequence of multi-sensor iterative fusion is optimized. Finally, the corresponding multi-sensor GM-PHD fusion tracking algorithm is constructed. To solve the problem that the multi-sensor adaptive order iterative fusion can not reflect the sensor quality gap, an Adaptive Iteratively Updating GM-PHD tracking algorithm PAIU-GM-PHD with weighted pseudo measurements is proposed. The simulation results show that, compared with the conventional multi-sensor iterative update GM-PHD tracking algorithm, the proposed algorithms can obtain more robust and accurate tracking results.
2022, 44(12): 4178-4184.
doi: 10.11999/JEIT210944
Abstract:
Armored target is currently great threatened by passive millimeter wave detection and guidance technology. An active cancellation millimeter-wave stealth method is proposed for enhancing the survivability of armored target in the future battlefield. By means of the low power noise from the millimeter wave jammer on armored target, the radiation temperature difference between target and background in practice is reduced. And then, the millimeter-wave radiometer in terminal-sensitive projectile could not detect and identify armored target so as to realize its passive stealth function. Compared to the traditional passive stealth methods based on shape and material, the proposed method can not only protect various types of targets under different backgrounds in action, but also has the advantages of strong mobility and simple engineering implementation. Finally, the experimental results demonstrate stealthy capability of armored target to Ka-band and W-band terminal-sensitive projectile radiometers above its 90° three-dimensional space can achieve –20～–8 dB and –15～–8 dB respectively, which it is also improved compared with the passive stealth methods.
Armored target is currently great threatened by passive millimeter wave detection and guidance technology. An active cancellation millimeter-wave stealth method is proposed for enhancing the survivability of armored target in the future battlefield. By means of the low power noise from the millimeter wave jammer on armored target, the radiation temperature difference between target and background in practice is reduced. And then, the millimeter-wave radiometer in terminal-sensitive projectile could not detect and identify armored target so as to realize its passive stealth function. Compared to the traditional passive stealth methods based on shape and material, the proposed method can not only protect various types of targets under different backgrounds in action, but also has the advantages of strong mobility and simple engineering implementation. Finally, the experimental results demonstrate stealthy capability of armored target to Ka-band and W-band terminal-sensitive projectile radiometers above its 90° three-dimensional space can achieve –20～–8 dB and –15～–8 dB respectively, which it is also improved compared with the passive stealth methods.
2022, 44(12): 4185-4193.
doi: 10.11999/JEIT210921
Abstract:
Focusing on how to improve the system spectral efficiency of Prolate Spheroidal Wave Functions(PSWFs) multi-carrier modulation system, a third constellation composed of additional constellation points on the basis of multi-carrier index modulation based on PSWFs with dual-mode method is introduced in this paper. The Multi-Carrier index Modulation based on PSWFs with Better multIple-Mode (BIM-MCM-PSWFs) is proposed. In this method, the signal index dimension is expanded and the number of modulation symbol combinations is increased through the multiple arrangement and combination of subcarriers in each sub block after grouping. The method proposed in this paper realizes the further utilization of spectrum resources in the multi-carrier index modulation based on PSWFs with dual-mode method, and improves effectively system spectral efficiency. Theoretical and simulation analysis show that, compared with the multi-carrier index modulation based on PSWFs with dual-mode method, the method proposed in this paper has a higher system spectral efficiency at the cost of appropriately sacrificing bit error performance. When n=9, k=1, m=4, Spectral Efficiency(SE) is increased by 20.1% at the expense of 0.70 dB of Bit Error Rate(BER).
Focusing on how to improve the system spectral efficiency of Prolate Spheroidal Wave Functions(PSWFs) multi-carrier modulation system, a third constellation composed of additional constellation points on the basis of multi-carrier index modulation based on PSWFs with dual-mode method is introduced in this paper. The Multi-Carrier index Modulation based on PSWFs with Better multIple-Mode (BIM-MCM-PSWFs) is proposed. In this method, the signal index dimension is expanded and the number of modulation symbol combinations is increased through the multiple arrangement and combination of subcarriers in each sub block after grouping. The method proposed in this paper realizes the further utilization of spectrum resources in the multi-carrier index modulation based on PSWFs with dual-mode method, and improves effectively system spectral efficiency. Theoretical and simulation analysis show that, compared with the multi-carrier index modulation based on PSWFs with dual-mode method, the method proposed in this paper has a higher system spectral efficiency at the cost of appropriately sacrificing bit error performance. When n=9, k=1, m=4, Spectral Efficiency(SE) is increased by 20.1% at the expense of 0.70 dB of Bit Error Rate(BER).
2022, 44(12): 4194-4203.
doi: 10.11999/JEIT211145
Abstract:
Focusing on the problem that the free space propagation model could only describe the propagation characteristics of vortex channel carrying Orbital Angular Momentum (OAM) in free space scenario, and the deterministic sparse multipath vortex channel model is strictly dependent on the propagation environment and could not accurately describe the propagation characteristics of OAM channel in real multipath scenario, a statistical modeling method for millimeter wave OAM multipath channel is proposed in this paper. In indoor corridor environment, Uniform Circular Array (UCA)-based OAM radiation transmission system is constructed, and the OAM multipath channel model is established based on optical ray theory and UCA radiation characteristics. Results show that the wavefront phase and amplitude of OAM channel in indoor corridor multipath environment can be accurately characterized by uniform distribution and Nakagami-m distribution in millimeter bands. The channel amplitude follows Rayleigh distribution when the propagation distance is large under the Line-of-Sight (LoS) and Non-LoS (NLoS) propagation conditions, and that follows Rician distribution when the propagation distance is small under the LoS propagation conditions.
Focusing on the problem that the free space propagation model could only describe the propagation characteristics of vortex channel carrying Orbital Angular Momentum (OAM) in free space scenario, and the deterministic sparse multipath vortex channel model is strictly dependent on the propagation environment and could not accurately describe the propagation characteristics of OAM channel in real multipath scenario, a statistical modeling method for millimeter wave OAM multipath channel is proposed in this paper. In indoor corridor environment, Uniform Circular Array (UCA)-based OAM radiation transmission system is constructed, and the OAM multipath channel model is established based on optical ray theory and UCA radiation characteristics. Results show that the wavefront phase and amplitude of OAM channel in indoor corridor multipath environment can be accurately characterized by uniform distribution and Nakagami-m distribution in millimeter bands. The channel amplitude follows Rayleigh distribution when the propagation distance is large under the Line-of-Sight (LoS) and Non-LoS (NLoS) propagation conditions, and that follows Rician distribution when the propagation distance is small under the LoS propagation conditions.
2022, 44(12): 4204-4210.
doi: 10.11999/JEIT211074
Abstract:
The architecture of Deep Neural Network (DNN) based detectors can affect the Symbol Detection (SD) accuracy and computational complexity. However, most of the works ignore the architecture selection method when establishing a DNN-based symbol detector. Moreover, the existing DNN detectors use complex architectures and only perform single-type modulated symbols detection. The Symbol Error Rate (SER) based strategy is proposed to design a low complexity Sparse AutoEncoder Detector (SAED) to tackle this problem. Furthermore, a Cumulant and Moment Feature Vector (CMFV)-based method is introduced for mixed symbols detection. Also, the designed symbol detector does not rely on a comprehensive knowledge of channel models and parameters but has the capability to detect various modulation signals. Simulation results show that the SER performance of the SAE symbol detector is close to the values of the Maximum Likelihood (ML) detection approach and provides a stable performance against phase offsets, frequency offsets, and under a limited training dataset.
The architecture of Deep Neural Network (DNN) based detectors can affect the Symbol Detection (SD) accuracy and computational complexity. However, most of the works ignore the architecture selection method when establishing a DNN-based symbol detector. Moreover, the existing DNN detectors use complex architectures and only perform single-type modulated symbols detection. The Symbol Error Rate (SER) based strategy is proposed to design a low complexity Sparse AutoEncoder Detector (SAED) to tackle this problem. Furthermore, a Cumulant and Moment Feature Vector (CMFV)-based method is introduced for mixed symbols detection. Also, the designed symbol detector does not rely on a comprehensive knowledge of channel models and parameters but has the capability to detect various modulation signals. Simulation results show that the SER performance of the SAE symbol detector is close to the values of the Maximum Likelihood (ML) detection approach and provides a stable performance against phase offsets, frequency offsets, and under a limited training dataset.
2022, 44(12): 4211-4217.
doi: 10.11999/JEIT220649
Abstract:
Considering the problems of high mobility and abnormal nodes in Unmanned Aerial Vehicle (UAV) communication, a Deep reinforcement learning based Trusted Geographic Routing protocol (DTGR) is proposed. A trusted third party is introduced to provide the trust of nodes. The difference between theoretical delay and real delay, and packet loss ratio are used as evaluation factors of trust degree. Routing selection is modeled as the Markov Decision Process (MDP). The state are constructed based on the neighbor nodes’ geographic location, the trust degree and the topology information. Then the routing decision can be output through the Deep Q Network(DQN). The action-value is adjusted by combining trust in reward function, to guide nodes to select the optimal next-hop. The simulation results show that DTGR has a lower average end-to-end delay and higher packet delivery ratio compared with existing schemes in UAV Ad hoc NETwork (UANET) with abnormal nodes. Besides, DTGR can effectively implement route selection and ensure network performance when the number or proportion of abnormal nodes changes.
Considering the problems of high mobility and abnormal nodes in Unmanned Aerial Vehicle (UAV) communication, a Deep reinforcement learning based Trusted Geographic Routing protocol (DTGR) is proposed. A trusted third party is introduced to provide the trust of nodes. The difference between theoretical delay and real delay, and packet loss ratio are used as evaluation factors of trust degree. Routing selection is modeled as the Markov Decision Process (MDP). The state are constructed based on the neighbor nodes’ geographic location, the trust degree and the topology information. Then the routing decision can be output through the Deep Q Network(DQN). The action-value is adjusted by combining trust in reward function, to guide nodes to select the optimal next-hop. The simulation results show that DTGR has a lower average end-to-end delay and higher packet delivery ratio compared with existing schemes in UAV Ad hoc NETwork (UANET) with abnormal nodes. Besides, DTGR can effectively implement route selection and ensure network performance when the number or proportion of abnormal nodes changes.
2022, 44(12): 4218-4228.
doi: 10.11999/JEIT210958
Abstract:
In ultra dense heterogeneous wireless networks, a vertical handoff algorithm considering Load Balancing and User Experience (LBUE) is proposed to solve the problem of network congestion caused by large-scale mobile terminals clustering in short time in urban traffic peak. Firstly, the network environment perception model is introduced to predict the future congestion degree of the network, and a network architecture integrating self-organizing network is proposed to alleviate network congestion. Secondly, the business fitness and negative return factor are defined, and an adaptive handoff decision algorithm based on Rank Sum Ratio(RSR) is proposed to screen out the most satisfactory target network for users in the current environment. Experimental results show that the algorithm can effectively reduce the blocking rate and call drop rate of terminal access network, achieve load balancing between networks and improve user experience.
In ultra dense heterogeneous wireless networks, a vertical handoff algorithm considering Load Balancing and User Experience (LBUE) is proposed to solve the problem of network congestion caused by large-scale mobile terminals clustering in short time in urban traffic peak. Firstly, the network environment perception model is introduced to predict the future congestion degree of the network, and a network architecture integrating self-organizing network is proposed to alleviate network congestion. Secondly, the business fitness and negative return factor are defined, and an adaptive handoff decision algorithm based on Rank Sum Ratio(RSR) is proposed to screen out the most satisfactory target network for users in the current environment. Experimental results show that the algorithm can effectively reduce the blocking rate and call drop rate of terminal access network, achieve load balancing between networks and improve user experience.
2022, 44(12): 4229-4237.
doi: 10.11999/JEIT211103
Abstract:
In order to solve the problem that the self-interference is difficult to eliminate in the transceiver system of the repeater jammer, a self-interference cancellation algorithm based on Deep Neural Network(DNN) is presented in this paper. This proposed algorithm can effectively eliminate the self-interference signal when the self-interference signal is correlated and mixed with the target signal. On this basis, the interference generation method of segmenting interception is used in this paper, which verifies the feasibility of jamming signal cancellation in the simultaneous transceiver system. It realizes the construction of radar jamming signal considering the pulse, which can quickly respond to enemy radars and occupy a favorable position in electronic countermeasures. In this paper, a typical Linear Frequency Modulation(LFM) and Binary Phase Shift Keying (BPSK) radar signal are used to generate a data set to train the DNN network, and the test set is used to test the output model of the network. The experimental results show that the self-interference cancellation algorithm based on DNN can effectively eliminate the self-interference signal when the target signal and self-interference signal are mixed in the simultaneous transceiver system, and the cancellation ratio can reach more than 26 dB when the signal to interference ratio is –8 dB.
In order to solve the problem that the self-interference is difficult to eliminate in the transceiver system of the repeater jammer, a self-interference cancellation algorithm based on Deep Neural Network(DNN) is presented in this paper. This proposed algorithm can effectively eliminate the self-interference signal when the self-interference signal is correlated and mixed with the target signal. On this basis, the interference generation method of segmenting interception is used in this paper, which verifies the feasibility of jamming signal cancellation in the simultaneous transceiver system. It realizes the construction of radar jamming signal considering the pulse, which can quickly respond to enemy radars and occupy a favorable position in electronic countermeasures. In this paper, a typical Linear Frequency Modulation(LFM) and Binary Phase Shift Keying (BPSK) radar signal are used to generate a data set to train the DNN network, and the test set is used to test the output model of the network. The experimental results show that the self-interference cancellation algorithm based on DNN can effectively eliminate the self-interference signal when the target signal and self-interference signal are mixed in the simultaneous transceiver system, and the cancellation ratio can reach more than 26 dB when the signal to interference ratio is –8 dB.
2022, 44(12): 4238-4245.
doi: 10.11999/JEIT211021
Abstract:
With the development of the Internet of Things, the demand for timeliness of information is increasing, and the freshness of information is becoming crucial. In order to maintain the freshness of information, the transmission scenario of multiple devices and single Mobile Edge Computing (MEC) server is studied in the joint system of Non-Orthogonal Multiple Access (NOMA) and MEC. In this scenario, how to allocate the amount of unload tasks and unload power to minimize the average update cost is a challenging problem. Considering the channel state variation in reality, an optimal unloading factor and unloading power decision are proposed based on Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm. Simulation results show that partial unloading can effectively reduce the average update cost, and MADDPG algorithm can further optimize the unloading power. By comparison, MADDPG algorithm is better than other schemes in reducing the average update cost, and the appropriate reduction of the number of equipment is better in reducing the average update cost.
With the development of the Internet of Things, the demand for timeliness of information is increasing, and the freshness of information is becoming crucial. In order to maintain the freshness of information, the transmission scenario of multiple devices and single Mobile Edge Computing (MEC) server is studied in the joint system of Non-Orthogonal Multiple Access (NOMA) and MEC. In this scenario, how to allocate the amount of unload tasks and unload power to minimize the average update cost is a challenging problem. Considering the channel state variation in reality, an optimal unloading factor and unloading power decision are proposed based on Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm. Simulation results show that partial unloading can effectively reduce the average update cost, and MADDPG algorithm can further optimize the unloading power. By comparison, MADDPG algorithm is better than other schemes in reducing the average update cost, and the appropriate reduction of the number of equipment is better in reducing the average update cost.
2022, 44(12): 4246-4255.
doi: 10.11999/JEIT211019
Abstract:
The existing indoor visible light positioning systems focus more on the positioning accuracy improvement and ignores the system's decoding success rate and the applicability of the positioning algorithm. The specific aspects are as follows: (1) Captured images with the blurring effect result in a lower decoding recognition rate; (2) Most existing systems use a single algorithm to achieve the positioning, and the dual-Light Emitting Diode (dual-LED) positioning algorithm exists the uncertainty for the rotation angle. The above problems result in the positioning accuracy decreasing or even the positioning system failure. First, the decoding algorithm based on the fringe width ratio is proposed to remove the defect of the artificial hard-threshold. Then, the joint positioning algorithm based on optimized rotation angle is proposed, which uses the smartphone orientation sensor to calibrate the rotation angle. Furthermore, a simple navigation application is designed. Experimental results show that the designed algorithms raise the decoding recognition rate up to 99% within 1.5 m, and the average positioning error is 3.998 cm.
The existing indoor visible light positioning systems focus more on the positioning accuracy improvement and ignores the system's decoding success rate and the applicability of the positioning algorithm. The specific aspects are as follows: (1) Captured images with the blurring effect result in a lower decoding recognition rate; (2) Most existing systems use a single algorithm to achieve the positioning, and the dual-Light Emitting Diode (dual-LED) positioning algorithm exists the uncertainty for the rotation angle. The above problems result in the positioning accuracy decreasing or even the positioning system failure. First, the decoding algorithm based on the fringe width ratio is proposed to remove the defect of the artificial hard-threshold. Then, the joint positioning algorithm based on optimized rotation angle is proposed, which uses the smartphone orientation sensor to calibrate the rotation angle. Furthermore, a simple navigation application is designed. Experimental results show that the designed algorithms raise the decoding recognition rate up to 99% within 1.5 m, and the average positioning error is 3.998 cm.
2022, 44(12): 4256-4264.
doi: 10.11999/JEIT211045
Abstract:
In the edge computing environment, the edge nodes have various kinds, wide distribution range and great differences in working environment, which makes the data security guarantee become very complicated. It is difficult to evaluate the credibility of the data transmitted by nodes effectively, thus affecting the security of the entire edge computing environment. To solve the problem, a novel collaborative trustworthy framework for edge computing is presented in the paper. By using the advantage of trusted computing technology, the concept of Trusted Computing Base for Edge computing (ETCB) is introduced, algorithms of the constructing ETCB are given and a trusted data validation protocol for the edge computing is designed. Meanwhile, the correctness of the algorithms and the security analysis of the protocol are also given. The framework can realize the distributed trusted verification of data without changing the original edge computing mode.
In the edge computing environment, the edge nodes have various kinds, wide distribution range and great differences in working environment, which makes the data security guarantee become very complicated. It is difficult to evaluate the credibility of the data transmitted by nodes effectively, thus affecting the security of the entire edge computing environment. To solve the problem, a novel collaborative trustworthy framework for edge computing is presented in the paper. By using the advantage of trusted computing technology, the concept of Trusted Computing Base for Edge computing (ETCB) is introduced, algorithms of the constructing ETCB are given and a trusted data validation protocol for the edge computing is designed. Meanwhile, the correctness of the algorithms and the security analysis of the protocol are also given. The framework can realize the distributed trusted verification of data without changing the original edge computing mode.
2022, 44(12): 4265-4273.
doi: 10.11999/JEIT211029
Abstract:
To address the problems of channel interference and inadequate utilization of spectrum resources in Wireless Mesh Networks (WMN) in the context of emergency communications, an Improved Discrete Bat Algorithm (IDBA) is proposed for solving the optimal Partially Overlapped Channels (POCs) assignment scheme. K-means clustering algorithm is used to optimize the network topology, the chaining behavior of the sea squirt is introduced to improve the local search capability, and a linear programming model with the goal of minimizing link-weighted interference is established to solve the network bottleneck link problem that may be caused in the case of traffic convergence. Results show that the method has a faster convergence speed and better search capability than other group intelligence optimization algorithm-based channel assignment methods at different network sizes. In addition, the method can significantly reduce the global interference and maintain the network stability when the nodes are dense.
To address the problems of channel interference and inadequate utilization of spectrum resources in Wireless Mesh Networks (WMN) in the context of emergency communications, an Improved Discrete Bat Algorithm (IDBA) is proposed for solving the optimal Partially Overlapped Channels (POCs) assignment scheme. K-means clustering algorithm is used to optimize the network topology, the chaining behavior of the sea squirt is introduced to improve the local search capability, and a linear programming model with the goal of minimizing link-weighted interference is established to solve the network bottleneck link problem that may be caused in the case of traffic convergence. Results show that the method has a faster convergence speed and better search capability than other group intelligence optimization algorithm-based channel assignment methods at different network sizes. In addition, the method can significantly reduce the global interference and maintain the network stability when the nodes are dense.
2022, 44(12): 4274-4283.
doi: 10.11999/JEIT211049
Abstract:
With the pervasiveness of secure data transmission techniques and increasing requirements of information authentication, the public key-based digital signature scheme has been extensively used in various fields. However, the process speed of digital signature has gradually become the bottleneck of various security and high-concurrency applications. In this paper, a high-throughput SM2 digital signature computing scheme based on Graphics Processing Unit(GPU) platform is proposed. Firstly, the basic operations are optimized by low-level instructions of GPU. Then, according to the characteristics of GPU platform, the addition chain of SM2 recommended prime number is reduced and the speed of modular inverse operation based on Fermat's theorem is improved. Furthermore, a pre-computing table is constructed and the repeated doubling algorithm is introduced to accelerate the unknown point multiplication. Due to the construction of pre-computing table, divergence of threads can be successfully avoided. The experiments show that the proposed scheme can effectively speed up SM2 algorithm, and the throughput of signing and verification can respectively reach 76.09 million ops and 3.46 million ops on RTX3090.
With the pervasiveness of secure data transmission techniques and increasing requirements of information authentication, the public key-based digital signature scheme has been extensively used in various fields. However, the process speed of digital signature has gradually become the bottleneck of various security and high-concurrency applications. In this paper, a high-throughput SM2 digital signature computing scheme based on Graphics Processing Unit(GPU) platform is proposed. Firstly, the basic operations are optimized by low-level instructions of GPU. Then, according to the characteristics of GPU platform, the addition chain of SM2 recommended prime number is reduced and the speed of modular inverse operation based on Fermat's theorem is improved. Furthermore, a pre-computing table is constructed and the repeated doubling algorithm is introduced to accelerate the unknown point multiplication. Due to the construction of pre-computing table, divergence of threads can be successfully avoided. The experiments show that the proposed scheme can effectively speed up SM2 algorithm, and the throughput of signing and verification can respectively reach 76.09 million ops and 3.46 million ops on RTX3090.
2022, 44(12): 4284-4293.
doi: 10.11999/JEIT211143
Abstract:
For improving the resolution of quantum decrypted image and computation complexity under the premise of ensuring quantum image cryptography algorithm security, an approach to quantum image chaotic encryption scheme based on Arnold transforms is proposed. In the paper, the chaotic signals generated by quantum cellular neural network are applied to control quantum Arnold transforms, quantum SWAP and quantum Controlled NOT (CNOT) operations which are utilized to process the plain quantum image to obtain the corresponding cipher image. Theoretical analyses show that the advantages of high security, fine resolution of decrypted images and considerable computation complexity are all presented in the proposed quantum gray image encryption scheme.
For improving the resolution of quantum decrypted image and computation complexity under the premise of ensuring quantum image cryptography algorithm security, an approach to quantum image chaotic encryption scheme based on Arnold transforms is proposed. In the paper, the chaotic signals generated by quantum cellular neural network are applied to control quantum Arnold transforms, quantum SWAP and quantum Controlled NOT (CNOT) operations which are utilized to process the plain quantum image to obtain the corresponding cipher image. Theoretical analyses show that the advantages of high security, fine resolution of decrypted images and considerable computation complexity are all presented in the proposed quantum gray image encryption scheme.
2022, 44(12): 4294-4303.
doi: 10.11999/JEIT211137
Abstract:
At present, network security and privacy have attracted extensive attention. Forward security is a security attribute of Authenticated Key Agreement protocol (AKA) proposed by Günther in 1989. Since then, this property has become one of the hot topics. This paper analyzes the security properties of two AKA protocols, MZK20 and VSR20. First, based on heuristic analysis and BAN logic, MZK20 protocol is proved that it does not satisfy weak forward security. Second, using heuristic analysis and Scyther, it is proved that VSR20 protocol does not fulfill forward security. Finally, the enhanced VSR20 protocol is designed and proved more secure than VSR20. The security of the modified VSR20 is verified both by the use of security reduction under eCK model and Scyther.
At present, network security and privacy have attracted extensive attention. Forward security is a security attribute of Authenticated Key Agreement protocol (AKA) proposed by Günther in 1989. Since then, this property has become one of the hot topics. This paper analyzes the security properties of two AKA protocols, MZK20 and VSR20. First, based on heuristic analysis and BAN logic, MZK20 protocol is proved that it does not satisfy weak forward security. Second, using heuristic analysis and Scyther, it is proved that VSR20 protocol does not fulfill forward security. Finally, the enhanced VSR20 protocol is designed and proved more secure than VSR20. The security of the modified VSR20 is verified both by the use of security reduction under eCK model and Scyther.
2022, 44(12): 4304-4311.
doi: 10.11999/JEIT210950
Abstract:
Based on orthogonal matrices, constructions of two Zero Correlation Zone (ZCZ) Aperiodic Complementary Sequence Sets (ZACSS) are proposed through different matrix transformation methods. Under the condition that the order of the orthogonal matrices can be evenly divided by the length of the zero-correlation zone, the parameters of obtained sequence sets are optimal, and the length of the ZCZ can be chosen flexibly. The sequence sets are constructed by the first method have ideal autocorrelation complementarity, and by further grouping, a set of intra-group complementary sequence sets can be obtained. A large number of optimal ZACSS can be constructed by different kinds of initial matrices and orthogonal matrices. The resultant sequence sets proposed in this paper can be applied to Multi-Carrier Code Division Multiple Access (MC-CDMA) systems as user address codes to eliminate multipath interference and multiple access interference.
Based on orthogonal matrices, constructions of two Zero Correlation Zone (ZCZ) Aperiodic Complementary Sequence Sets (ZACSS) are proposed through different matrix transformation methods. Under the condition that the order of the orthogonal matrices can be evenly divided by the length of the zero-correlation zone, the parameters of obtained sequence sets are optimal, and the length of the ZCZ can be chosen flexibly. The sequence sets are constructed by the first method have ideal autocorrelation complementarity, and by further grouping, a set of intra-group complementary sequence sets can be obtained. A large number of optimal ZACSS can be constructed by different kinds of initial matrices and orthogonal matrices. The resultant sequence sets proposed in this paper can be applied to Multi-Carrier Code Division Multiple Access (MC-CDMA) systems as user address codes to eliminate multipath interference and multiple access interference.
2022, 44(12): 4312-4318.
doi: 10.11999/JEIT211104
Abstract:
Few-weight linear codes have important applications in constructing authentication codes, association schemes and secret sharing schemes. How to construct few-weight linear codes has always been an important topic of coding theory. In this paper, irreducible quasi-cyclic codes of index 2 over finite fields are constructed by selecting a special defining set. The weight distribution of several classes of irreducible quasi-cyclic codes of index 2 are determined by using Gaussian periods over finite fields. Some classes of 2-weight linear codes and 3-weight linear codes are obtained. The results show that two of the three classes of 2-weight linear codes constructed in this paper are Maximum Distance Separable (MDS) codes and the other class reaches Griesmer bound.
Few-weight linear codes have important applications in constructing authentication codes, association schemes and secret sharing schemes. How to construct few-weight linear codes has always been an important topic of coding theory. In this paper, irreducible quasi-cyclic codes of index 2 over finite fields are constructed by selecting a special defining set. The weight distribution of several classes of irreducible quasi-cyclic codes of index 2 are determined by using Gaussian periods over finite fields. Some classes of 2-weight linear codes and 3-weight linear codes are obtained. The results show that two of the three classes of 2-weight linear codes constructed in this paper are Maximum Distance Separable (MDS) codes and the other class reaches Griesmer bound.
2022, 44(12): 4319-4327.
doi: 10.11999/JEIT211066
Abstract:
Timed-Release Encryption (TRE) is a cryptographic primitive called "sending messages into the future", the receiver can not decrypt the ciphertext until a designed time in the future. Currently, most TRE schemes use a non-interactive single time server approach, where the system user is able to decrypt properly, relying on a time trapdoor calculated and broadcast by a time server at the designed decryption time. If the single time server is attacked or corrupted, it is prone to directly threaten the TRE security application. Therefore, a single time server needs to be “distributed” into multiple ones. However, existing multiple time servers TRE schemes do not provide neither the security analysis nor the strict formal security proofs. To deal with this problem, a new Multiple Time Servers TRE (MTSTRE) scheme based on Bilinear Diffie-Hellman (BDH) problem in the random oracle model is proposed, a concrete scheme with provable security and a general scheme is proposed, and then it is strictly proved that the concrete scheme is security under adaptive chosen-plaintext attack. Efficiency analysis shows that compared with the most effective existing multiple time servers TRE scheme, the calculation efficiency of the concrete scheme is slightly improved.
Timed-Release Encryption (TRE) is a cryptographic primitive called "sending messages into the future", the receiver can not decrypt the ciphertext until a designed time in the future. Currently, most TRE schemes use a non-interactive single time server approach, where the system user is able to decrypt properly, relying on a time trapdoor calculated and broadcast by a time server at the designed decryption time. If the single time server is attacked or corrupted, it is prone to directly threaten the TRE security application. Therefore, a single time server needs to be “distributed” into multiple ones. However, existing multiple time servers TRE schemes do not provide neither the security analysis nor the strict formal security proofs. To deal with this problem, a new Multiple Time Servers TRE (MTSTRE) scheme based on Bilinear Diffie-Hellman (BDH) problem in the random oracle model is proposed, a concrete scheme with provable security and a general scheme is proposed, and then it is strictly proved that the concrete scheme is security under adaptive chosen-plaintext attack. Efficiency analysis shows that compared with the most effective existing multiple time servers TRE scheme, the calculation efficiency of the concrete scheme is slightly improved.
2022, 44(12): 4328-4336.
doi: 10.11999/JEIT211169
Abstract:
The nonlinear functions introduced into the existing multi-scroll attractor chaotic systems are mostly step function, saturation function, multi-logic level pulse function and so on. Therefore, the system hardware complexity increases with the increase of the number of scrolls, which makes its hardware implementation more difficult. To solve this problem, a Logistic level pulse function is designed and a new chaotic system with multi-scroll based on the Lorenz system is constructed by the nonautonomous pulse control method. The system’s dynamic characteristics are analyzed and the system is implemented based on FPGA chip. Finally, the application of the system in image encryption is given. The analysis results show that the hardware complexity of the system designed in this paper is independent of the number of scrolls, so the FPGA circuit can produce different multi-scroll attractors only by changing the control parameters without the change of the RTL codes; Compared with the Lorenz system, the designed multi-scroll system has more sensitive parameters so as to have larger key space and resist brute-force attack more effectively when applied to image encryption.
The nonlinear functions introduced into the existing multi-scroll attractor chaotic systems are mostly step function, saturation function, multi-logic level pulse function and so on. Therefore, the system hardware complexity increases with the increase of the number of scrolls, which makes its hardware implementation more difficult. To solve this problem, a Logistic level pulse function is designed and a new chaotic system with multi-scroll based on the Lorenz system is constructed by the nonautonomous pulse control method. The system’s dynamic characteristics are analyzed and the system is implemented based on FPGA chip. Finally, the application of the system in image encryption is given. The analysis results show that the hardware complexity of the system designed in this paper is independent of the number of scrolls, so the FPGA circuit can produce different multi-scroll attractors only by changing the control parameters without the change of the RTL codes; Compared with the Lorenz system, the designed multi-scroll system has more sensitive parameters so as to have larger key space and resist brute-force attack more effectively when applied to image encryption.
2022, 44(12): 4337-4347.
doi: 10.11999/JEIT211054
Abstract:
In order to solve the problems of low embedding rate of reversible data hiding in encrypted domain with multi-users and weak disaster tolerance of cover image, a reversible data hiding in encrypted images based on polynomial secret sharing scheme is proposed. The cover image is divided into multiple shadow images and stored to multiple users, so that the image disaster tolerance can be enhanced. Separability is achieved by combining two embedding algorithms, where Algorithm 1 embeds additional data into the redundancy coefficients of the polynomial during the process of image sharing to generate shadow images containing additional data, which can extract data after image reconstruction; Algorithm 2, for any shadow image, embeds additional data by using the additive homomorphism of secret sharing, which supports extracting additional data directly from the shadow image. Experiments are performed in different threshold and shadow image compression rates. When the compression rate is 50%, the embedding rate reaches 4.18 bit per pixel (bpp) for the (3, 4) threshold scheme and 3.78 bpp for the (3, 5) threshold scheme. The results show that the algorithms support extracting additional data from the shadow images and the reconstructed images respectively, which achieves the separability of the scheme. The proposed scheme has a higher embedding rate and lower computational complexity than the existing schemes, and is more practical.
In order to solve the problems of low embedding rate of reversible data hiding in encrypted domain with multi-users and weak disaster tolerance of cover image, a reversible data hiding in encrypted images based on polynomial secret sharing scheme is proposed. The cover image is divided into multiple shadow images and stored to multiple users, so that the image disaster tolerance can be enhanced. Separability is achieved by combining two embedding algorithms, where Algorithm 1 embeds additional data into the redundancy coefficients of the polynomial during the process of image sharing to generate shadow images containing additional data, which can extract data after image reconstruction; Algorithm 2, for any shadow image, embeds additional data by using the additive homomorphism of secret sharing, which supports extracting additional data directly from the shadow image. Experiments are performed in different threshold and shadow image compression rates. When the compression rate is 50%, the embedding rate reaches 4.18 bit per pixel (bpp) for the (3, 4) threshold scheme and 3.78 bpp for the (3, 5) threshold scheme. The results show that the algorithms support extracting additional data from the shadow images and the reconstructed images respectively, which achieves the separability of the scheme. The proposed scheme has a higher embedding rate and lower computational complexity than the existing schemes, and is more practical.
2022, 44(12): 4348-4356.
doi: 10.11999/JEIT210971
Abstract:
Medical cloud storage service is one of the most significant applications in cloud computing. Simultaneously, the integrity of outsourced medical data and users’ identity privacy-preservation have been more and more important. To this end, an outsourced cloud storage medical data lightweight integrity verification scheme is proposed for wireless medical sensor networks, supporting conditional identity anonymity. The scheme combines the homomorphic hash function to design an aggregated signature to enable a Third Party Auditor (TPA) to check the integrity of outsourced medical data effectively. The scheme stores auditing auxiliary information on TPA side and uses the homomorphic property of the homomorphic hash function to optimize the calculations on TPA side to a constant, which reduces greatly the computational costs of TPA. The scheme enables TPA to perform batch verification on multiple data files, and the verification costs are nearly constant, independent of the number of data files. In addition, this scheme prevents effectively TPA from recovering the original medical data by solving the linear equations, and a conditional identity anonymous algorithm is designed, thus the Private Key Generator (PKG) could generate the anonymous identity of a user and corresponding singing key. Even if the attacker intercepts the medical data transmitted by the user, it can not know the real identity of the data. In addition, the complex certificates management is efficiently avoided, and PKG could also trace and revoke the real identities of misbehaved users efficiently. The security analysis and performance evaluation demonstrate that this scheme could be securely and efficiently deployed in wireless medical sensor networks.
Medical cloud storage service is one of the most significant applications in cloud computing. Simultaneously, the integrity of outsourced medical data and users’ identity privacy-preservation have been more and more important. To this end, an outsourced cloud storage medical data lightweight integrity verification scheme is proposed for wireless medical sensor networks, supporting conditional identity anonymity. The scheme combines the homomorphic hash function to design an aggregated signature to enable a Third Party Auditor (TPA) to check the integrity of outsourced medical data effectively. The scheme stores auditing auxiliary information on TPA side and uses the homomorphic property of the homomorphic hash function to optimize the calculations on TPA side to a constant, which reduces greatly the computational costs of TPA. The scheme enables TPA to perform batch verification on multiple data files, and the verification costs are nearly constant, independent of the number of data files. In addition, this scheme prevents effectively TPA from recovering the original medical data by solving the linear equations, and a conditional identity anonymous algorithm is designed, thus the Private Key Generator (PKG) could generate the anonymous identity of a user and corresponding singing key. Even if the attacker intercepts the medical data transmitted by the user, it can not know the real identity of the data. In addition, the complex certificates management is efficiently avoided, and PKG could also trace and revoke the real identities of misbehaved users efficiently. The security analysis and performance evaluation demonstrate that this scheme could be securely and efficiently deployed in wireless medical sensor networks.
2022, 44(12): 4357-4366.
doi: 10.11999/JEIT211010
Abstract:
Three Dimensional-High Efficiency Video Coding (3D-HEVC) standard is the latest Three-Dimensional (3D) video coding standard, but the coding complexity increases greatly due to the introduction of depth map coding technology. Among them, the quad-tree partition of depth map intra-frame Coding Unit (CU) accounts for more than 90% of the coding complexity in 3D-HEVC. Therefore, for the intra-frame coding of depth map in 3D-HEVC, considering the high complexity of CU quad-tree partition, a fast prediction scheme of CU partition structure based on deep learning is proposed. Firstly, the dataset of CU partition structure information for learning depth map is constructed. Secondly, a Multi-Branch Convolutional Neural Network (MB-CNN) model for predicting the CU partition structure is built. Then, the MB-CNN model is trained by using the built dataset. Finally, the MB-CNN model is embedded into the 3D-HEVC test platform, which reduces greatly the complexity of CU partition by predicting the partition structure of CU in depth map intra-frame coding. Experimental results show that the proposed algorithm reduces effectively the coding complexity of 3D-HEVC without significant synthesized view quality distortion. Specifically, compared to the standard method, the coding complexity on the standard test sequence is reduced by 37.4%.
Three Dimensional-High Efficiency Video Coding (3D-HEVC) standard is the latest Three-Dimensional (3D) video coding standard, but the coding complexity increases greatly due to the introduction of depth map coding technology. Among them, the quad-tree partition of depth map intra-frame Coding Unit (CU) accounts for more than 90% of the coding complexity in 3D-HEVC. Therefore, for the intra-frame coding of depth map in 3D-HEVC, considering the high complexity of CU quad-tree partition, a fast prediction scheme of CU partition structure based on deep learning is proposed. Firstly, the dataset of CU partition structure information for learning depth map is constructed. Secondly, a Multi-Branch Convolutional Neural Network (MB-CNN) model for predicting the CU partition structure is built. Then, the MB-CNN model is trained by using the built dataset. Finally, the MB-CNN model is embedded into the 3D-HEVC test platform, which reduces greatly the complexity of CU partition by predicting the partition structure of CU in depth map intra-frame coding. Experimental results show that the proposed algorithm reduces effectively the coding complexity of 3D-HEVC without significant synthesized view quality distortion. Specifically, compared to the standard method, the coding complexity on the standard test sequence is reduced by 37.4%.
2022, 44(12): 4367-4375.
doi: 10.11999/JEIT211081
Abstract:
Hive's existing audit function can not make compliance judgment on the purpose of data operation. To solve the above problems, a Hive data operation compliance analysis method based on subgraph isomorphism is proposed. Firstly, the modeling method of Hive data operation and compliance rules based on graph is proposed to form data traceability graph and compliance rule graph; Then, the compliance judgment of data operation is modeled as the matching problem of traceability graph and compliance graph, and a solution algorithm based on subgraph isomorphism is proposed. Finally, the experimental verification is carried out in the data governance platforms Apache Atlas and Hive. The experimental results show that the proposed method has higher compliance verification efficiency than the collection based, VF2 and Ullmann compliance verification.
Hive's existing audit function can not make compliance judgment on the purpose of data operation. To solve the above problems, a Hive data operation compliance analysis method based on subgraph isomorphism is proposed. Firstly, the modeling method of Hive data operation and compliance rules based on graph is proposed to form data traceability graph and compliance rule graph; Then, the compliance judgment of data operation is modeled as the matching problem of traceability graph and compliance graph, and a solution algorithm based on subgraph isomorphism is proposed. Finally, the experimental verification is carried out in the data governance platforms Apache Atlas and Hive. The experimental results show that the proposed method has higher compliance verification efficiency than the collection based, VF2 and Ullmann compliance verification.
2022, 44(12): 4376-4383.
doi: 10.11999/JEIT211034
Abstract:
In order to solve the problems that traditional knowledge reasoning methods can not combine computing power and interpretability, and it is difficult to learn quickly in few-shot scenarios, a Model-Agnostic Meta-Learning (MAML) reasoning framework is proposed in this paper, which combines semantic path and Bidirectional Encoder Representations for Transformers (BERT), and consists of two stages: base-training and meta-training. In base-training stage, the graph reasoning instances is represented by semantic path and BERT model, which is used to calculate the link probability and save reasoning experience offline by fine-tuning. In meta-training stage, the gradient meta-information based on the base-training process of multiple relations is obtained by this framework, which realizes the initial weight optimization, and completes the rapid learning of knowledge under few-shot. Experiments show that better performance in link prediction and fact prediction can be achieved by the base-training reasoning framework, and fast convergence of some few-shot reasoning problems can be achieved by the meta-learning framework.
In order to solve the problems that traditional knowledge reasoning methods can not combine computing power and interpretability, and it is difficult to learn quickly in few-shot scenarios, a Model-Agnostic Meta-Learning (MAML) reasoning framework is proposed in this paper, which combines semantic path and Bidirectional Encoder Representations for Transformers (BERT), and consists of two stages: base-training and meta-training. In base-training stage, the graph reasoning instances is represented by semantic path and BERT model, which is used to calculate the link probability and save reasoning experience offline by fine-tuning. In meta-training stage, the gradient meta-information based on the base-training process of multiple relations is obtained by this framework, which realizes the initial weight optimization, and completes the rapid learning of knowledge under few-shot. Experiments show that better performance in link prediction and fact prediction can be achieved by the base-training reasoning framework, and fast convergence of some few-shot reasoning problems can be achieved by the meta-learning framework.
2022, 44(12): 4384-4394.
doi: 10.11999/JEIT210946
Abstract:
In order to reduce the co-adaptability interference of the original feature to the Non-negative Matrix Factorization (NMF) algorithm and improve the performance of non-negative matrix factorization subspace learning and clustering performance, a novel multi-constrained semi-supervised non-negative matrix factorization algorithm based on Sinkhorn distance feature scaling is proposed. First, the algorithm is feature-scaled by the Sinkhorn distance to the original input matrix to improve the correlation between features of the same type of data in the space, then, the dual graph manifold structure combined with the sample label information and the norm sparsity constraint are embedded in the model as a dual regular term, so that the decomposed base matrix has sparse characteristics and strong spatial expression ability. Finally, the objective function of the proposed algorithm is optimized by Karush-Kuhn-Tucker (KKT) conditions, and effective multiplication update rules are obtained. Through the comparative analysis of the results of multiple clustering experiments on multiple image data sets and translational noise data, the algorithm proposed in this paper has a strong subspace learning ability and is more robust to translational noise.
In order to reduce the co-adaptability interference of the original feature to the Non-negative Matrix Factorization (NMF) algorithm and improve the performance of non-negative matrix factorization subspace learning and clustering performance, a novel multi-constrained semi-supervised non-negative matrix factorization algorithm based on Sinkhorn distance feature scaling is proposed. First, the algorithm is feature-scaled by the Sinkhorn distance to the original input matrix to improve the correlation between features of the same type of data in the space, then, the dual graph manifold structure combined with the sample label information and the norm sparsity constraint are embedded in the model as a dual regular term, so that the decomposed base matrix has sparse characteristics and strong spatial expression ability. Finally, the objective function of the proposed algorithm is optimized by Karush-Kuhn-Tucker (KKT) conditions, and effective multiplication update rules are obtained. Through the comparative analysis of the results of multiple clustering experiments on multiple image data sets and translational noise data, the algorithm proposed in this paper has a strong subspace learning ability and is more robust to translational noise.
2022, 44(12): 4395-4404.
doi: 10.11999/JEIT211101
Abstract:
With the continuous development of video acquisition equipment and technology, the number of videos has grown rapidly. It is a challenging task in video retrieval to find target video moments accurately in massive videos. Cross-modal video moment retrieval is to find a moment matching the query from the video database. Existing works focus mostly on matching the text with the moment, while ignoring the context content in the adjacent moment. As a result, there exists the problem of insufficient expression of feature relation. In this paper, a novel moment retrieval network is proposed, which highlights the significant features through residual channel attention. At the same time, a temporal adjacent network is designed to capture the context information of the adjacent moment. Experimental results show that the proposed method achieves better performance than the mainstream candidate matching based and video-text features relation based methods.
With the continuous development of video acquisition equipment and technology, the number of videos has grown rapidly. It is a challenging task in video retrieval to find target video moments accurately in massive videos. Cross-modal video moment retrieval is to find a moment matching the query from the video database. Existing works focus mostly on matching the text with the moment, while ignoring the context content in the adjacent moment. As a result, there exists the problem of insufficient expression of feature relation. In this paper, a novel moment retrieval network is proposed, which highlights the significant features through residual channel attention. At the same time, a temporal adjacent network is designed to capture the context information of the adjacent moment. Experimental results show that the proposed method achieves better performance than the mainstream candidate matching based and video-text features relation based methods.
2022, 44(12): 4405-4415.
doi: 10.11999/JEIT211112
Abstract:
Infrared and visible image fusion algorithms based on deep learning rely on artificially designed similarity functions to measure the similarity between input and output. The unsupervised learning method can not effectively utilize the ability of neural networks to extract deep features, resulting in unsatisfactory fusion results. Considering this problem, a new fusion degradation model of infrared and visible image is proposed in this paper, which regards infrared and visible images as the degraded images produced by ideal fusion images through mixed degradation processes. Secondly, a data enhancement scheme for simulating image degradation is proposed, and a large number of simulated degradation images are generated by using high-definition datasets for training the network. Finally, a simple and efficient end-to-end network model and its network training framework are designed based on the proposed degradation model. The experimental results show that the method proposed in this paper not only has good visual effects and performance indicators, but also can effectively suppress interferences such as illumination, smoke and noise.
Infrared and visible image fusion algorithms based on deep learning rely on artificially designed similarity functions to measure the similarity between input and output. The unsupervised learning method can not effectively utilize the ability of neural networks to extract deep features, resulting in unsatisfactory fusion results. Considering this problem, a new fusion degradation model of infrared and visible image is proposed in this paper, which regards infrared and visible images as the degraded images produced by ideal fusion images through mixed degradation processes. Secondly, a data enhancement scheme for simulating image degradation is proposed, and a large number of simulated degradation images are generated by using high-definition datasets for training the network. Finally, a simple and efficient end-to-end network model and its network training framework are designed based on the proposed degradation model. The experimental results show that the method proposed in this paper not only has good visual effects and performance indicators, but also can effectively suppress interferences such as illumination, smoke and noise.
2022, 44(12): 4416-4425.
doi: 10.11999/JEIT211157
Abstract:
With the rapid development of aerospace electronic technology, especially the wide use of Commercial Off-The-Shelf(COTS), the space computing capacity has increased significantly. Space edge computing organizes the widely dispersed space-based computing resources into a distributed and collaborative cloud environment through inter-satellite links, so as to realize resource complementarity and task collaboration, which can effectively get rid of the dependence on the ground and improve the speed of service response. Firstly, the requirements development status, existing problems and challenges of space edge computing are clarified. Then, a space edge computing architecture is proposed, which is described from the perspectives of physical architecture, functional architecture, software architecture and service process. Finally, the key technologies are summarized and analyzed in order to provide valuable suggestions and references for follow-up research.
With the rapid development of aerospace electronic technology, especially the wide use of Commercial Off-The-Shelf(COTS), the space computing capacity has increased significantly. Space edge computing organizes the widely dispersed space-based computing resources into a distributed and collaborative cloud environment through inter-satellite links, so as to realize resource complementarity and task collaboration, which can effectively get rid of the dependence on the ground and improve the speed of service response. Firstly, the requirements development status, existing problems and challenges of space edge computing are clarified. Then, a space edge computing architecture is proposed, which is described from the perspectives of physical architecture, functional architecture, software architecture and service process. Finally, the key technologies are summarized and analyzed in order to provide valuable suggestions and references for follow-up research.
2022, 44(12): 4426-4440.
doi: 10.11999/JEIT210972
Abstract:
The rapid development of depth sensor and laser scanning technology allows people to collect easily a large amount of point cloud data. Point cloud data can provide rich scene and object information, and has become the preferred research object for applications such as autonomous driving, virtual reality, and robot navigation. As an important research method, point cloud segmentation has received extensive attention from industry and academia. Especially driven by deep learning, the effect of point cloud segmentation has been significantly improved. In order to stimulate future research, the latest progress in point cloud segmentation are comprehensively reviewed , and comparative studies are conducted from the perspective of indirect and direct processing of point clouds. Among them, the method based on indirect processing can be divided into the multi-view and voxel-based method. The method based on direct processing can be divided into point processing, optimized convolutional neural networks , graph convolution, timing and unsupervised learning. Then the basic ideas and characteristics of the representative methods are introduced in each category. In addition, the common data sets and evaluation indexes of point cloud segmentation are sorted out. Finally, the future of point cloud classification and segmentation technology is prospected.
The rapid development of depth sensor and laser scanning technology allows people to collect easily a large amount of point cloud data. Point cloud data can provide rich scene and object information, and has become the preferred research object for applications such as autonomous driving, virtual reality, and robot navigation. As an important research method, point cloud segmentation has received extensive attention from industry and academia. Especially driven by deep learning, the effect of point cloud segmentation has been significantly improved. In order to stimulate future research, the latest progress in point cloud segmentation are comprehensively reviewed , and comparative studies are conducted from the perspective of indirect and direct processing of point clouds. Among them, the method based on indirect processing can be divided into the multi-view and voxel-based method. The method based on direct processing can be divided into point processing, optimized convolutional neural networks , graph convolution, timing and unsupervised learning. Then the basic ideas and characteristics of the representative methods are introduced in each category. In addition, the common data sets and evaluation indexes of point cloud segmentation are sorted out. Finally, the future of point cloud classification and segmentation technology is prospected.
