关闭×
Office
Quick Search
Adv Search
Current Issue
30 August 2021, Volume 48 Issue 4
  • Edge offloading strategy for the multi-base station game in ultra-dense networks
    WANG Ruyan,WU He,CUI Yaping,WU Dapeng,ZHANG Hong
    2021, 48(4):  1-10.  doi:10.19665/j.issn1001-2400.2021.04.001
    Abstract ( 87 )   HTML( 61 )   PDF (1919KB) ( 61 )   Save

    Mobile edge computing provides powerful computing capabilities for the wireless network to enrich the user experience.However,in the current mobile edge computing network,the problems of small coverage density and hotspot overload of the central node should be skillfully overcome.The combination of the ultra-dense network and mobile edge computing can provide a feasible solution for addressing the above problems.A distributed edge computing architecture for ultra-dense networks is designed,and a multi-base station game offloading algorithm is proposed to minimize the system overhead.In the proposed algorithm,the lagrange multiplier method is used to solve the problem of computing resource allocation,and then the matching game theory is exploited to obtain the optimal offloading strategy,so that the mutual benefits of both users and mobile edge computing servers are maximized.Simulation results show that compared with the random and greedy offloading algorithms,the proposed algorithm achieves a significant reduction in the system overhead,with the average overhead saving being up to 28.66%.

    Minimum cost of node deployment strategy for heterogeneous sensor networks
    LI Ming,HU Jiangping,CAO Xiaoli
    2021, 48(4):  11-19.  doi:10.19665/j.issn1001-2400.2021.04.002
    Abstract ( 42 )   HTML( 21 )   PDF (2929KB) ( 21 )   Save

    A minimum-cost-deployment strategy based on objective optimization is proposed to address the problems existing in heterogeneous sensor networks such as high density of the nodes,bad target coverage and connectivity performances,and high deployment spending.The problem we aim to solve is characterized by different parameters in terms of the deployment cost of positions and the cost of the sensors.The enhanced version of coral reef optimization (CRO for short) algorithm is utilized to solve the problem of how to select the proper positions and sensors to achieve the minimum deployment cost of heterogeneous sensor networks which can fulfill both k-coverage and m-connectivity requirements.The enhanced version of the CRO is named ECRO.In the ECRO two methods are employed to improve the optimization efficiency of the CRO.One method is that inspired by the process of the harmony search algorithm the operators such as HMCR and PAR of the harmony search algorithm blend with the CRO.The other one is that the excellent solutions are reused to exploit the knowledge and experience accumulated in the process of running the CRO.For comparison purposes,a greedy algorithm is also proposed.Simulation experiments show that compared with some related existing algorithms,the proposed algorithm reduces the deployment cost of sensor nodes while fulfilling the requirements of k-coverage and m-connectivity requirements.

    Method for distributed deployment of the virtual network function manager for MEC
    MA Yue,ZHANG Yumei
    2021, 48(4):  20-26.  doi:10.19665/j.issn1001-2400.2021.04.003
    Abstract ( 35 )   HTML( 12 )   PDF (1529KB) ( 12 )   Save

    Aiming at achieving efficient virtual network function deployment under the Mobile edge computing (MEC) architecture,a virtual network function manager (VNFM) deployment method based on the immune optimization algorithm is proposed.First,a mixed integer programming model is used to build the mathematical model of VNFM deployment.Then,a deployment solution based on the immune optimization algorithm is given.Aiming at obtaining the optimal deployment solution towards the minimum communication cost,the algorithm comprehensively considers the antibody affinity and antigen affinity of chromosomes with respect to individual fitness in the population and the diversity characteristics of the immune system,respectively.Simulation results show that compared with the current deployment method,the individual evaluation mechanism of the proposed algorithm can more effectively evaluate the fitness and similarity of individuals in VNFM deployment problems.The proposed method can effectively prevent the algorithm from falling into the local optimum during the deployment process.The optimal solution improves the performance of the algorithm;it can speed up the algorithm’s convergence efficiency and simultaneously reduce the CPU time for algorithm execution.

    Algorithm for multivariate similarity localization based on dynamic fuzzy matching
    QIN Ningning,WU Yisong,YANG Le
    2021, 48(4):  27-35.  doi:10.19665/j.issn1001-2400.2021.04.004
    Abstract ( 32 )   HTML( 13 )   PDF (2360KB) ( 13 )   Save

    Considering the problem that the traditional area division method in fingerprint positioning cannot guarantee the accuracy of area matching,this paper proposes a fuzzy matching algorithm.Its matching range can be dynamically adjusted,and while improving the positioning speed,the algorithm can decrease the defects of mismatch caused by area division.In the position calculation stage,the algorithm uses multivariate similarity coefficients and introduces environmental parameters to redistribute the weights of neighboring points.This algorithm overcomes the adverse effects of fluctuation in environment and equipment,and reduces positioning errors.Tested by an actual scene,on the premise of ensuring the accuracy of matching,the proposed algorithm reduces the amount of fingerprint matching by more than 60% compared with global matching,and experiment verifies that the positioning accuracy is improved by more than 17%,compared with the existing algorithms.

    Fast measurement of the reflection coefficient of the plate medium in the time domain
    GONG Minghao,WEI Bing,HU Mingwei,WANG Weidong
    2021, 48(4):  36-41.  doi:10.19665/j.issn1001-2400.2021.04.005
    Abstract ( 23 )   HTML( 8 )   PDF (1965KB) ( 8 )   Save

    In order to solve the problems faced in the frequency domain measurement of the broadband reflection characteristics of materials in practical engineering,this paper proposes a fast test scheme for free-space ultra wideband reflection characteristics based on the time domain method.Through the construction of the time domain measurement system,the analysis of the extraction method of the pulse signal based on the "time window",the calculation of the reflection coefficient,etc.,the system advantages are comprehensively analyzed,and the actual measurement plan is formed.Measured results show that the scheme has the advantages of low measurement environment requirements,fast measurement speed,great convenience and high accuracy,and that the reflection coefficient of plate-like materials in the common radar band (2~18 GHz) can be obtained by one measurement.

    Simulation methods for inductive pulsed current injection on the coaxial cable
    CUI Zhitong,WEI Bing,WU Wei,SUN Beiyun
    2021, 48(4):  42-49.  doi:10.19665/j.issn1001-2400.2021.04.006
    Abstract ( 15 )   HTML( 8 )   PDF (1750KB) ( 8 )   Save

    This paper aims at a quantitative analysis and the design of the inductive PCI (Pulsed Current Injection)on coaxial cables.First the induced current on the cable shield is derived by Spice circuit modeling.And the reasonable measurement method and coupling equations between the shield and inner conductor are proposed under the specific boundary conditions decided by the PCI.Then a Spice+Matlab simulation method for the inductive PCI on coaxial cables is developed.Furthermore,the negative inductor and the fitting circuit model of transfer impedance are expressed in Spice based on the principle of circuit equivalent response.Then the lumped approximate model of the coupling circuit between the shield and inner conductor is developed by subdividing the cable into several electrical small subsections.Finally a complete Spice model for the PCI on coaxial cables is constructed.By numerical and experimental validation,the two aforesaid methods have an appreciable agreement and show a high simulation precision and convergence in the time domain.Both of them will be very helpful to analysis and design of the PCI on coaxial cables.

    Influence of the second-order ionospheric delay on GPS PPP time transfer
    DING Shuo,YANG Haiyan,YANG Xuhai,ZHANG Zhe,ZHAO Ku...
    2021, 48(4):  50-56.  doi:10.19665/j.issn1001-2400.2021.04.007
    Abstract ( 22 )   HTML( 10 )   PDF (2358KB) ( 10 )   Save

    In view of the practical problem that the second-order ionospheric delay error (Ion2) is not corrected in Precise Point Positioning (PPP) time transfer,the PPP observation equation with the Ion2 is derived.The PPP clocks and time transfer with and without the Ion2 correction are compared and analyzed using two test periods.Based on the measured data,analysis shows that the influence of the Ion2 on the PPP clocks at a low-latitude are more significant than that of mid-latitude and polar regions.The difference series of PPP clock solutions shows fixed deviation plus some diurnal variations,and the sign of fixed deviation is opposite in the northern and southern hemispheres.The maximum influence of the Ion2 on the long time-links transfer can be up to more than 26ps,but for the short time-links,it can be ignored.Therefore the Ion2 correction is necessary for sub-nanosecond PPP time transfer over long time-links,especially when the time-links are made by the stations located in low-latitude regions.

    Optimization algorithm for ultra-sparse arrays in regional energy focusing
    YANG Zhongping,ZHOU Qingsong,ZHANG Jianyun
    2021, 48(4):  57-63.  doi:10.19665/j.issn1001-2400.2021.04.008
    Abstract ( 16 )   HTML( 7 )   PDF (1318KB) ( 7 )   Save

    The setting of ultra-sparse arrays is one of key factors for the performance of regional energy focusing.This paper focuses on one of its typical applications:Precision Electronic Warfare by UAV,and proposes an optimization algorithm for ultra-sparse arrays to maximize the performance of regional energy focusing.First,we establish the joint optimization model of the interference waveform and the positions of elements.Second,particle swarm optimization (PSO) is adopted to iterate the positions of elements.Then,the target value of the subproblem is solved for the fitness criterion,as the current array is known in each step of the iteration.Finally,we get the optimal solution of the interference waveform and the ultra-sparse array by this method.Numerical results indicate that the algorithm can optimize the array effectively,and improve the indicators of regional energy focusing obviously.Whether the localization errors of elements exist or not,the algorithm provides a better performance than existing algorithms which only design the interference waveform.

    Short-arc association algorithm for joint observation of multiple star sensors
    HUANG Qiushi,ZHANG Yasheng,FENG Fei
    2021, 48(4):  64-72.  doi:10.19665/j.issn1001-2400.2021.04.009
    Abstract ( 16 )   HTML( 6 )   PDF (1562KB) ( 6 )   Save

    The star sensor is a high-precision attitude sensor with the ability to shoot space targets intermittently.It can be used as a new type of space target monitoring platform.Accurately associating the track segments of the star sensor with intermittent observations is the prerequisite for precise orbit determination of space targets.In order to solve the problem of trajectory association where the initial orbit cannot be determined because the target observation time is too short,a short arc association algorithm for space targets based on hypothetical boundary values is proposed.By using the multi-satellite sensor joint positioning method,only the angle measurement data of the star sensor are converted into the spatial coordinates of the target,and each point from the track segment to be associated is taken as a boundary.Based on the boundary association hypothesis,the theoretical values of the space target orbit under all assumed boundary conditions are calculated.The Mahalanobis distance and chi-square test are used to make related judgments.The algorithm in this paper can better adapt to many-to-many track association,and can reject uncorrelated tracks more accurately.After simulation,the algorithm in this paper is superior to the short arc correlation method based on the admissible region to determine the initial orbit and sine fitting in terms of association accuracy and execution time.

    Maximum likelihood registration for systemic error based on statistical linear regression
    LI Jiawei,JIANG Jing,WU Weihua,ZHENG Yujun
    2021, 48(4):  73-82.  doi:10.19665/j.issn1001-2400.2021.04.010
    Abstract ( 18 )   HTML( 9 )   PDF (1477KB) ( 9 )   Save

    Generally,there are non-random systemic errors in target detection with the cooperative multi-sensor system.In order to solve this problem,a maximum likelihood registration algorithm based on statistical linear regression (SLR-MLR) is presented.The registration equation for the multi-sensor system is established first by jointly maximizing the likelihood function of the target state and systemic error,on the basis of which the proposed algorithm utilizes a set of diverse regression points to handle the linearization problem of the nonlinear measurement transformation.The regression equation for the target state with respect to unbiased measurement is constructed through statistical linear regression,and then the first two statistical properties of the projected state can be obtained.Moreover,the algorithm uses the likelihood maximization iteration to seek the solution of the registration equation,thus achieving the joint estimation for the systemic error and target state.Simulation results show that the SLR-MLR can achieve the registration of multiple sensors in each observation dimension,and has a higher accuracy and near computational complexity compared with the classical MLR.

    Design of the broadband microstrip array antenna for blast furnace surface monitoring
    WANG Yu,YU Jingna,CHEN Xianzhong,HOU Qingwen
    2021, 48(4):  83-90.  doi:10.19665/j.issn1001-2400.2021.04.011
    Abstract ( 17 )   HTML( 4 )   PDF (1600KB) ( 4 )   Save

    Aiming at the requirements of high-resolution imaging and high integration of the blast furnace radar,this paper presents a method for designing a wideband microstrip array antenna,which broadens the antenna bandwidth through the design of parasitic patches and air layers.By combining multi-input multi-output (MIMO) radar and synthetic aperture radar (SAR) imaging principles,a linear MIMO array is designed,and a near-field simulation imaging experiment is performed on the simulated feed line through the wave number domain imaging algorithm.Simulation results show that the gain of the main lobe of the antenna can reach 14.05 dBi,the reflection coefficient is less than -10 dB,the absolute bandwidth is 5.25 GHz,the operating frequency is 20.67~25.92 GHz,and the range resolution is increased to 3 cm compared with the existing blast furnace radar.The average error of the azimuth direction of the simulated material line imaging is 0.008 m,and the range direction is 0.0011 m.Compared with the traditional microstrip array antenna,this antenna effectively widens the bandwidth,and the range resolution is higher than that of the traditional blast furnace radar.It can accurately obtain the shape information on the simulated material line,and has an engineering application value for blast furnace surface monitoring.

    Design of a low temperature coefficient bandgap reference with a wide temperature range
    AN Jinghui,WU Chenjian
    2021, 48(4):  91-96.  doi:10.19665/j.issn1001-2400.2021.04.012
    Abstract ( 13 )   HTML( 4 )   PDF (921KB) ( 4 )   Save

    In order to meet the requirements of different applications and markets for the accuracy and reliability of IoT chips,a low temperature coefficient bandgap reference with a wide temperature range is proposed.On the basis of the traditional Banba bandgap reference structure,the circuit utilizes high-order temperature compensation technology and piecewise temperature compensation technology to improve the curvature of the output reference voltage.The temperature coefficient of the circuit is reduced.At the same time,the operating temperature range of the circuit is extended.The circuit performances are verified in the TSMC 180 nm CMOS process.Test results show that the temperature coefficient of the circuit is as low as 7.2×10-6/℃ in the range of-40 ℃ to 160 ℃.The power supply rejection ratio at a low frequency is -48.52 dB.The static current under the 1.8 V power supply voltage is 68.38 μA,and the core area of the chip is 0.025 mm2.

    Multi-channel charge readout ASIC for X-ray detection
    LIU Hong,XU Le,TIAN Tong
    2021, 48(4):  97-102.  doi:10.19665/j.issn1001-2400.2021.04.013
    Abstract ( 12 )   HTML( 4 )   PDF (1118KB) ( 4 )   Save

    To solve the problem of large area and high power consumption of the traditional multi-channel charge readout circuit for X-ray detection,a new multi-channel charge readout circuit structure is presented.The circuit integrates 32 charge acquisition channels,including an integrator,a differential output buffer,a digital control logic circuit and other functional modules.The proposed structure reduces the power consumption and area by integrating the sample and hold circuit and the correlated double sampling circuit into the output buffer.In order to reduce the offset and low-frequency noise,the correlated double sampling technology has been adopted.In addition,the way of overlapping the outputs between the even and odd channels is adopted to realize continuous transmission of the multi-channel signals.The chip has been fabricated using the CMOS 0.25 μm process with the chip size of 2.87 mm×2.68 mm.Measurement results show that the power consumption of each channel is 1.5 mW under the 5.0 V voltage supply and 2.5 V reference voltage.When the integrated capacitor is 7.8 pF,the integration non-linearity is less than 0.1% and the dynamic range reaches 13 100.

    Method for inpainting blind images using the game and L0 constraint
    FENG Xiangchu,WANG Ping,HE Ruiqiang
    2021, 48(4):  103-112.  doi:10.19665/j.issn1001-2400.2021.04.014
    Abstract ( 24 )   HTML( 7 )   PDF (1509KB) ( 7 )   Save

    Image inpainting is the process of restoring the original image from the observed image with missing pixels using the prior information on the original image.Most image inpainting models assume that the missing areas of the image are known.However,inpractical applications,the information on these missing areas is difficult to obtain directly.In order to solve this problem,a new image inpainting model is established by using the sparse priori of L0 norm and game theory.The new model is suitable for the two cases of known and unknown image missing areas.According to the structure of the objective function,an effective proximal alternating direction method of multipliers and a game-based alternating framework are proposed to solve the corresponding minimization problem,and the convergence of the model under certain conditions is analyzed.Compared with the existing inpainting models,numerical experiments show that the models and algorithms proposed can lead to better results and robustness insubjective and objective quality evaluation than the image inpainting methods available.

    General reliability model of the non-MDS coded storage system
    NIE Shiqiang,ZHENG Xuda,LIU Zhaohua,WU Weiguo,DONG...
    2021, 48(4):  113-119.  doi:10.19665/j.issn1001-2400.2021.04.015
    Abstract ( 16 )   HTML( 8 )   PDF (986KB) ( 8 )   Save

    An open problem is how different configurations influence the reliability of a storage system using non-maximum distance separable codes as redundancy strategy.This paper proposes a repairable probability algorithm for solving data objects with non-maximum distance separable code encoding by considering the construction matrix of non-maximum distance separable codes.This algorithm exhaustively loses all possible combinations of several blocks and judges whether the matrix corresponding to each combination is reversible for calculating the probability of recoverability.We propose an analytical model based on the Markov chain to quantify the reliability of the non-maximum distance separable coded storage system.This model could quantify the impact of a series of design factors on the reliability of the storage system,such as the effect of non-maximum distance separable code configuration,the capacity of the storage system,the capacity of the object-based storage device nodes,the repair bandwidth,the mean time to data loss of the object-based storage device nodes and so on.Finally,the numerical analytical method is used to verify the correctness of the model and the influence of different factors on the reliability of the storage system.Our model enables system practitioners to decide the appropriate configuration based on their reliability requirements.

    Adaptive simulated annealing particle swarm optimization algorithm
    YAN Qunmin,MA Ruiqing,MA Yongxiang,WANG Junjie
    2021, 48(4):  120-127.  doi:10.19665/j.issn1001-2400.2021.04.016
    Abstract ( 32 )   HTML( 10 )   PDF (1377KB) ( 10 )   Save

    Particle swarm optimization is widely used in various fields because of the few parameters to be set and the simple calculation structure.In order to improve the optimization speed and accuracy of the PSO,and to avoid falling into the local optimal solution,an adaptive simulated annealing PSO is proposed,which uses the hyperbolic tangent function to control the inertia weight factor for nonlinear adaptive changes,uses linear change strategies to control 2 learning factors,introduces the simulation annealing operation,set a temperature according to the initial state of the population,guide the population to accept the difference solution with a certain probability according to the Metropolis criterion,and ensure the ability to jump out of the local optimal solution.To verify the effect of the algorithm proposed in this paper,7 typical test functions and 5 algorithms proposed in the literature are selected for comparison and testing.According to the average value,standard deviation and number of iterations of the optimization results,the algorithm proposed in this paper has greatly improved the iteration accuracy,convergence speed and stability so as to overcome the shortcomings of particle swarm optimization.

    Image denoising and boundary extraction based on game theory
    QIAO Yu,FENG Xiangchu
    2021, 48(4):  128-135.  doi:10.19665/j.issn1001-2400.2021.04.017
    Abstract ( 17 )   HTML( 4 )   PDF (2724KB) ( 4 )   Save

    Half-quadratic regularization is a classical image denoising method.In removing image noise,the image boundary can be obtained.Since the boundary obtained by the half-quadratic regularization model is too fuzzy and the denoising effect is not ideal,the half-quadratic regularization model is improved by the game method,the image is denoised and the boundary is extracted simultaneously.Two participants are defined,with the classical half-quadratic regularization method used as the target function of denoising,and a relatively novel global sparse gradient model selected as the target function of boundary extraction.The two participants,image denoising and boundary extraction,iterate alternately in a game process,with their convergence points as the Nash equilibrium points.The proposed model is applied to various types of images,and the algorithm proposed can lead to good results in both numerical results and visual effects.Experimental results show that the proposed algorithm can effectively improve the half-quadratic regularization model,thus obtaining better denoising and boundary extraction effects.

    DB-SMOTE and multi-layer stacking used for Arrhythmia recognition
    WANG Bo,DENG Ke
    2021, 48(4):  136-143.  doi:10.19665/j.issn1001-2400.2021.04.018
    Abstract ( 13 )   HTML( 3 )   PDF (2012KB) ( 3 )   Save

    In order to improve the recognition performance of the electrocardiogram,especially the recognition accuracy of minor diseases,this paper proposes the electrocardiogram recognition architecture based on the DB-SMOTE algorithm and multi-layered stacking model.The DB-SMOTE algorithm is proposed to solve the problem because the classical oversampling SMOTE algorithm ignores the intra-class unbalance of minority-class data.The new algorithm utilizes DBSCAN clustering to divide the data of minority classes into multiple clusters and filter out the noise samples,takes the boundary data of each cluster as the main body to generate new samples,and analyzes visually by tSNE.The performance of a single classifier cannot meet the requirements,so a multi-layered stack classification is used for identification,which is divided into two parts:the first is based on KNN,Xgboost and GBDT,and the feature F is mapped to F';the second part of the model is to identify the feature F'.This architecture has a 99.66% accuracy rate in identifying the electrocardiogram and can improve the recognition accuracy of minor diseases well,so it can be used to identify arrhythmias effectively.

    Gait recognition method based on spatial-temporal convolution
    XU Huanhuan,LI Hongmei,LI Fuyu,SUN Xuemei
    2021, 48(4):  144-150.  doi:10.19665/j.issn1001-2400.2021.04.019
    Abstract ( 20 )   HTML( 8 )   PDF (1428KB) ( 8 )   Save

    As a unique biological characteristic that can be recognized from a distance,gait has a wide application prospect in the fields of identity identification,public security and medical diagnosis.However,the accuracy of gait recognition will be affected by external factors,such as the shooting angle,the pedestrian’s wearing and the change in the status of carrying a bag.Based on the above problems,this paper puts forward the gait recognition method based on spatial-temporal convolution,which uses the convolution neural network to extract gait features,and employs the repeated extraction of the gait features of adjacent frames to make up for the missing information,so that more spatial-temporal information can be obtained.Finally,the proposed method is validated on the CASIA-B dataset.Experimental results show that the proposed method can improve the rank-1 accuracy when a pedestrian walks normally,carrying a bag and wearing a coat.

    Interframe target regression network for vehicle detection in UAV video
    ZHANG Zhi,ZHENG Jin
    2021, 48(4):  151-158.  doi:10.19665/j.issn1001-2400.2021.04.020
    Abstract ( 25 )   HTML( 14 )   PDF (2030KB) ( 14 )   Save

    UAV video has many advantages of flexible view,continuous view and wide monitoring scope,and at the same time,there are many problems,such as crowded targets,strong motion noises and so on,which make target detection difficult.To solve these problems,this paper proposes a video vehicle detection algorithm based on the interframe target regression network.According to the characteristics of crowded vehicles in UAV video,soft non maximum suppression is proposed as the detecting-box merging strategy of FCOS,and thus a single-frame vehicle detector is constructed.In order to deal with the problem that the single-frame detector can be easily disturbed by motion noise when it is directly applied to video detection,thus resulting in the change of the confidence level for the same target,an interframe target regression network is designed.The target features of adjacent multiple frames are fused by using interframe movement continuity,and the fused features are matched with the target features of the current frame to output the prediction results.Finally,the detection performance is improved by correcting prediction results through single-frame detection results.Compared with FCOS and FGFA,the average precision of the proposed algorithm is improved by 2% and 5% respectively,reaching 47.42%.Experimental results show that it is better than the existing FCOS and FGFA,and has better robustness and generalization.

    Potential emotion recognition based on the fusion of the spatio-temporal neural network and facial pulse signals
    SONG Jianqiao,WANG Feng,NIU Jin,SHI Zezhou,MA Junh...
    2021, 48(4):  159-167.  doi:10.19665/j.issn1001-2400.2021.04.021
    Abstract ( 13 )   HTML( 0 )   PDF (1490KB) ( 0 )   Save

    (Micro) expression has a certain effect on emotion recognition,but in the case of artificial concealment,it is prone to misjudgment.Although the recognition effect of physiological signals is more accurate,the data collection is often complicated,which is not convenient for rapid personnel emotion checking.In response to the above problem,this paper adopts a non-contact chromatic model-based method to collect pulse signals,extract features based on the pulse signals,and integrate the proposed spatio-temporal neural network to realize potential emotion recognition.Experimental results show that the proposed two-way latent emotion recognition framework can well integrate the emotion information contained in micro-expressions and physiological signals,and that the effect in micro-expression recognition is improved to some extent compared with the current micro-expression recognition algorithms commonly used at this stage.

    Detection of voice transformation spoofing using the dense convolutional neural network
    WANG Yong,SU Zhuoyi,ZHU Zhengyu
    2021, 48(4):  168-175.  doi:10.19665/j.issn1001-2400.2021.04.022
    Abstract ( 11 )   HTML( 0 )   PDF (1700KB) ( 0 )   Save

    Voice transformation (VT) spoofing refers to the operations for hiding the speaker’s identity which change a speaker’s acoustic features by speech processing algorithms and result in extremely high false reject rates for automatic speaker recognition (ASR) systems.VT spoofing is implemented with a low cost and has been integrated in many audio editing tools,thus presenting serious threats to social security.However,the research on VT spoofing detection is still insufficient.Hence,in this paper we propose a dense convolutional neural network (DenseNet) based VT detection method for distinguishing spoofed voices and genuine ones.The proposed network consists of 135 layers in total.By maximizing the skip-layers,the data transmission can be enhanced,and both the deep and shallow edge features can be used for classification,so as to alleviate the degradation phenomenon and further to improve detection accuracy.Experimental results show that the detection accuracy with various spoofing factors is over 98%.

    Unbalanced data weighted boundary point integration undersampling method
    HE Yunbin,LENG Xin,WAN Jing
    2021, 48(4):  176-183.  doi:10.19665/j.issn1001-2400.2021.04.023
    Abstract ( 12 )   HTML( 1 )   PDF (749KB) ( 1 )   Save

    In order to effectively solve the problem that boundary points are deleted directly from unbalanced data and effectively maintain the information on most kinds of data,a clustering-based weighted boundary point integration undersampling algorithm is proposed.First,the algorithm extracts the number of minority class sets as the initial number of clustering centers of majority class sets to cluster.Then,the variation coefficient is introduced to identify the boundary points,and the identified boundary points are weighted so that the weighted boundary points can be added to the unbalanced data processing.Then,the cluster density is used to divide majority class sets into the high-density cluster and low-density cluster,delete the low-density cluster,and finally obtain the reduced majority of the sample sets.Then,the reduced majority of class samples is combined with the minority of class samples to form a balanced data set,which is trained with the Ada boost to get the final classification model.This method can be used to reduce the dataset and improve the efficiency of execution.The results show that the proposed method can effectively handle the problem of unbalanced data,and improve the execution efficiency and accuracy of the under-sampling algorithm for unbalanced data weighted boundary point integration.

    Information system business affecting impact evaluation method
    XIE Lixia,BAI Yu
    2021, 48(4):  184-191.  doi:10.19665/j.issn1001-2400.2021.04.024
    Abstract ( 12 )   HTML( 1 )   PDF (940KB) ( 1 )   Save

    To effectively analyze the impact of continuous business interruption on the information system,an information system business affecting impact evaluation method (IBAIE) is proposed.First,we identify the main businesses of the system and quantify their vulnerability.Then,we take the businesses as the nodes of network topology and weight nodes based on the business significance.The orderly association between businesses is taken as the edges which are weighted based on the association between asset and business,and the directed weighted business network topology is obtained.Finally,the method of directed weighted network structure entropy is used to evaluate the change of system business network structure during the period from business interruption to recovery.Experiments show that this method has obvious advantages in evaluating accuracy compared to other methods,and can be applied in ensuring information system security.

    Correlation fault attack on AES
    WANG Xingxin,HU Wei,TAN Jing,ZHU Jiacheng,TANG Shi...
    2021, 48(4):  192-199.  doi:10.19665/j.issn1001-2400.2021.04.025
    Abstract ( 16 )   HTML( 5 )   PDF (2369KB) ( 5 )   Save

    Fault injection attack is an effective cryptanalysis method.However,most existing fault injection attacks have strict restrictions on the location,time and number of faults injected,require complicated mathematical derivation during the key recovery process or need a huge amount of time to train fault attack templates.This paper proposes a comprehensive correlation fault injection attack on AES implementations of different key lengths,leveraging the correlation in the fault effect propagation in AES to recover the key.Our attack method uses a more flexible fault model in terms of the location and number of fault injections while only requiring simple correlation analysis to recover the key.Experimental results using AES implementations of variable key sizes show that random faults injected at any position before the mix-columns operation in the-2 round will allow successful recovery of the last round key through correlation analysis of the fault effects at the inputs of the S-Box in the final round.Additional random faults injected at any position before the mix-columns operation in the-3 round will allow the recovery of the round key before the final round.The key search complexity of the proposed method is 216.Two correct and faulty ciphertext pairs or four faulty ciphertexts under the same plaintext are sufficient to recover the original key of AES-128 and four correct and faulty ciphertext pairs or eight faulty ciphertexts under the same plaintext are sufficient to recover the original key of AES-192 and AES-256.

    Method for stroke lesion segmentation using the primary-auxiliary path attention compensation network
    HUI Haisheng,ZHANG Xueying,WU Zelin,LI Fenglian
    2021, 48(4):  200-208.  doi:10.19665/j.issn1001-2400.2021.04.026
    Abstract ( 14 )   HTML( 3 )   PDF (1780KB) ( 3 )   Save

    When the feature of stroke lesions is non-distinct,and the boundary between the lesions and the healthy brain tissue is difficult to distinguish,the segmentation model based on the self-attention mechanism is prone to generate a wrong attention coefficient map of the focus area,which affects the segmentation performance.To solve this problem,based on the global-attention-upsample attention U-Net (GAU-A-UNet),we propose a primary-auxiliary path attention compensation network (PAPAC-Net).The primary path network is responsible for accurate lesion segmentation and outputting the segmentation results while the auxiliary path network generates a tolerant auxiliary attention compensation coefficient to compensate for the primary path network’s potential attention coefficient map errors.Two compound loss functions are also proposed to realize the different functions for the primary and auxiliary path networks.Experimental results show that our GAU-A-UNet and PAPAC-Net both have a significant improvement in segmentation performance,which proves the effectiveness of our method.

News
Download
Links