Current Issue

25 March 2023, Volume 44 Issue 3

Previous Issue    Next Issue

Papers

Method based on contrastive incremental learning for fine-grained malicious traffic classification

Yifeng WANG, Yuanbo GUO, Qingli CHEN, Chen FANG, Renhao LIN, Yongliang ZHOU, Jiali MA

2023, 44(3):  1-11.  doi:10.11959/j.issn.1000-436x.2023068

Asbtract ( 490 )   HTML ( 192)   PDF (2121KB) ( 614 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to protect against continuously emerging unknown threats, a new method based on contrastive incremental learning for fine-grained malicious traffic classification was proposed.The proposed method was based on variational auto-encoder (VAE) and extreme value theory (EVT), and the high accuracy could be achieved in known, few-shot and unknown malicious classes and new classes were also identified without using a large number of old task samples, which met the demand of storage and time cost in incremental learning scenarios.Specifically, the contrastive learning was integrated into the encoder of VAE, and the A-Softmax was used for known and few-shot malicious traffic classification, EVT and the decoder of VAE were used for unknown malicious traffic recognition, all classes could be recognized without a lot of old samples when learning new tasks by using VAE reconstruction and knowledge distillation methods.Experimental results indicate that the proposed method is efficient in known, few-shot and unknown malicious classes, and has greatly reduced the forgetting speed of old knowledge in incremental learning scenarios.

Adaptive resource optimization mechanism for blockchain sharding in digital twin edge network

Li JIANG, Shengli XIE, Hui TIAN

2023, 44(3):  12-23.  doi:10.11959/j.issn.1000-436x.2023044

Asbtract ( 447 )   HTML ( 169)   PDF (3186KB) ( 701 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To address the security and privacy issues for data sharing in digital twin edge networks, a new distributed and secure data sharing mechanism was developed by leveraging blockchain sharding.Considering dynamic and time varying features in digital twin edge networks and mapping errors of physic networks, an adaptive resource allocation model was formulated which jointly optimized cluster head selection, local base station (BS) consensus access selection, spectrum and computation consensus resources allocation, in order to maximize transactional throughput for the blockchain sharding.Moreover, a digital twin edge networks empowered two-layered proximal policy optimization (PPO) algorithm was designed to obtain the optimal solutions.Numerical results illustrate that the proposed algorithm can effectively improve the transactional throughput for the blockchain sharding.Meanwhile, it can well adapt to the mapping errors.

Research on channel-fixed delay error estimation of wideband digital array radar

Xiaobei ZHANG, Quan ZHANG, Ming FANG, Kun QIN, Qi ZHANG

2023, 44(3):  24-32.  doi:10.11959/j.issn.1000-436x.2023067

Asbtract ( 216 )   HTML ( 88)   PDF (1220KB) ( 430 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

A channel-fixed delay error estimation method was proposed to tackle the performance degradation of digital beamforming caused by the channel-fixed delay error of wideband digital array radar.First, the integer part of the fixed delay of the channel was estimated through the correlation integral peak of the baseband transmitted signal and the baseband received signal containing unknown delay parameters.Then, the fractional part of the fixed delay of the channel was estimated with parabolic interpolation by using the correlation peak and its adjacent two points.Finally, the multi-channel fixed delay error was calculated and compensated.The simulation results show that the multi-channel fixed delay estimation of wideband digital array radar can be realized on Xilinx xcku115.The estimative precision can reach 0.001～0.014 times the sampling gap when the signal-to-noise ratio of the input signal is 10～30 dB.The proposed method can meet the engineering application requirements and improve the performance of digital beamforming in wideband digital array radar.

Design, implementation and formal verification of BGP proxy for mimic router

Jin ZHANG, Qiang GE, Weihai XU, Yiming JIANG, Hailong MA, Hongtao YU

2023, 44(3):  33-44.  doi:10.11959/j.issn.1000-436x.2023065

Asbtract ( 241 )   HTML ( 45)   PDF (2589KB) ( 414 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To ensure the safety and correctness of the critical ‘mimic bracket’ components such as protocol proxies of mimic routers, a BGP (border gateway protocol) proxy was designed and implemented, and formal methods were applied to verify the safety and correctness of the BGP proxy.The BGP packets communicated between the peer routers and the master actor were monitored by the BGP proxy.The BGP sessions with the slave actors on behalf of peer routers were established, ensuring the consistency of the BGP protocol states for all actors.The formal specification of the BGP proxy was written based on separation logic.The VeriFast theorem prover was used to prove that the program had no memory safety problems such as null pointer reference.Furthermore, the formal verification of high-level attributes of each module in BGP proxy was also conducted to strictly ensure that the implementation met the specification.The implementation to proof code ratio of BGP proxy is about 1.8:1, and the implementation to proof labor hour ratio is about 1:3.The formally verified BGP proxy consume 0.16 seconds to process 100 000 BGP routes, which is about 7 times as long as the unverified one.Works done provide a reference for applying formal methods to verify the safety and correctness of critical components in mimic defense equipment and systems.

Security communication and energy efficiency optimization strategy in UAV-aided edge computing

Xueyong YU, Lixiang QIU, Jianing SONG, Hongbo ZHU

2023, 44(3):  45-54.  doi:10.11959/j.issn.1000-436x.2023032

Asbtract ( 483 )   HTML ( 135)   PDF (2429KB) ( 749 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The flexible mobility of the unmanned aerial vehicle (UAV) has attracted widespread attention in the mobile edge computing (MEC) system.However, the existence of eavesdroppers in the air makes it a huge challenge for its secure transmission.In order to solve the contradiction between high safe communication rate and low energy consumption, the concept of security communication energy efficiency was introduced, that was, the ratio between UAV safe communication transmission rate and UAV energy consumption.Firstly, to subject the task delay constraint, limited UAV CPU frequency and task offloading rate constraint, an offloading strategy was proposed to maximize the energy efficiency of secure communication by jointly optimizing the legal UAV hover location, CPU frequency allocation and distinguishing the complexity of computing tasks, while improving the security communication in the UAV-MEC scenario from the perspective of physical layer security.Secondly, to address the non-convex optimization problem, it was decomposed into three sub-problems that were solved with block coordinate descent and the successive convex approximation (SCA) methods respectively.The simulation results show that, with different task complexity, the proposed strategy can balance the relationship between the overall secure communication performance and energy consumption, while meeting the offloading requirements of ground terminals.And then it improves secrecy energy efficiency.

Optical quantum positioning method based on random resampling of scattering free path

Mu ZHOU, Jing ZHANG, Yong WANG, Xiaolong YANG, Lingxia LI

2023, 44(3):  55-65.  doi:10.11959/j.issn.1000-436x.2023053

Asbtract ( 158 )   HTML ( 21)   PDF (3286KB) ( 436 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To solve the problem that the quantum signal was disturbed by the scattering environment, which lead to large fluctuation of the positioning error and poor positioning robustness, an optical quantum positioning method based on random resampling of the scattering free path was proposed.The scattering model was established based on the Monte Carlo theory, and the scattering free path matrix was constructed by combining the environment information and coincidence counting results.The scattering free program was resampled, and the relative distance error of different light sources in different time groups was calculated.Based on that, the light sources were dynamically selected, and the position of the target object was estimated.Experimental results show that the positioning error of the proposed method can achieve the centimeter level, and the confidence probability of the positioning error within 0.12 m is 90.57%.

Blockchain network layer anomaly traffic detection method based on multiple classifier integration

Qianyi DAI, Bin ZHANG, Song GUO, Kaiyong XU

2023, 44(3):  66-80.  doi:10.11959/j.issn.1000-436x.2023066

Asbtract ( 315 )   HTML ( 52)   PDF (4623KB) ( 330 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To improve the comprehensive generalized feature perception ability of mixed attack traffic on the blockchain network layer, and enhance the performance of abnormal traffic detection, a blockchain layer traffic anomaly detection method was proposed that supported the comprehensive judgement of data anomaly with a strong generalisation capability.Firstly, to expand the difference of the input feature subset of the base classifier used, a feature subset selection algorithm based on discrimination degree and redundant information was proposed, and the output of high sensitivity subset terms was stimulated during the feature screening process, while the generation of redundant information was suppressed.Then, the stochastic variance reduction gradient algorithm was introduced into the bagging integration algorithm to realize the dynamic adjustment of the voting weights of each base modeland improve thecapability in detecting the generalised hybrid abnormal attack traffic.Finally, LBoF algorithm was proposed to map the low-dimensional numerical vector output by the integrated algorithm to a high-dimensional space.The discrepancy of data point spatial density distribution of various samples were amplified based on the potential difference between data points to increase the recall rate of anomalous data point detection.The experimental results show that in detecting multiple hybrid attack traffic on blockchain layers, the proposed method presents an increase in the anomaly detection accuracy and recall rate, which is 1.57% and 2.71%, respectively, compared with methods based on a single classifier integration.

Simultaneous vehicular location and velocity detection towards 6G integrated communication and sensing

Bingpeng ZHOU, Shanshan MA

2023, 44(3):  81-92.  doi:10.11959/j.issn.1000-436x.2023041

Asbtract ( 361 )   HTML ( 72)   PDF (1245KB) ( 731 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Joint vehicle location and velocity estimation algorithm was proposed for 6G millimeter wave integrated communication and sensing (ICAS), which was challenging due to random channel fading and multipath interference.A novel receiver state sensing (RSS) algorithm was proposed to simultaneously estimate vehicle (receiver) location and velocity, in conjunction with reflection channel estimation.Moreover, a novel multi-carrier Doppler calibration scheme was developed to reduce the disturbance of Doppler effect on RSS performance via frequency shift compensation.The impact of system parameters, e.g., signal bandwidth, the number of transceiver antenna and sub-carriers, on RSS performance was evaluated as well.It was verified by simulations that the proposed RSS algorithm outperformed state-of-the-art baseline methods due to the employed problem-specific system design.In addition, it was shown by simulations that the achieved RSS error was reduced with the increasing of sub-carrier frequency, the number of transceiver antennas and sub-carriers, and it was increased with the increasing distance between base station and vehicle receiver.Particularly, the vehicle location estimation accuracy was increasing with system bandwidth, but it was invariant with central carrier frequency, since only baseband features of received signal were employed for vehicle localization.In contrast, the vehicle velocity detection accuracy was increasing with central carrier frequency, yet it was insensitive to system bandwidth, given the fixed number of sub-carriers.

CHA: cross-chain based hierarchical architecture for practicable blockchain regulatory

Pujie JING, Liangmin WANG, Xuewen DONG, Yushu ZHANG, Qian WANG, Sohail Muhammad

2023, 44(3):  93-104.  doi:10.11959/j.issn.1000-436x.2023038

Asbtract ( 316 )   HTML ( 60)   PDF (2202KB) ( 698 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Researching the problems with blockchain regulation is crucial.In view of this, CHA (cross-chain based hierarchical architecture) was proposed to address the problems in blockchain regulatory, and a “chain of custody-business chain” cross-chain collaboration chain-based regulatory model was designed.The decentralized of information system regulatory was improved by building a decentralized governance chain, which realized regulatory by separating the strategy layer from the regulatory layer and the business layer, and achieved separation of regulatory behavior and regulatory technology, thus ensuring the versatility of the blockchain regulatory structure.Finally, the distributed feasibility of the regulatory framework is demonstrated through a formal analysis method, and then the technical feasibility of the scheme in terms of cross-chain regulatory reliability, general feasibility and scalability applicable to different regulatory services are proved.

Objective assessment of communication speech interference effect based on feature fusion

Yun LIN, Huaitao XU, Sen WANG, Sicheng ZHANG, Long ZHUANG

2023, 44(3):  105-116.  doi:10.11959/j.issn.1000-436x.2023043

Asbtract ( 209 )   HTML ( 31)   PDF (4580KB) ( 258 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In view of the objective assessment problem of the effect of communication speech interference, methods based on multi-measurements and multimodal fusion were proposed.First, the interfered speech was preprocessed by the endpoint detection algorithm and time warping algorithm.Then, the content of speech was extracted and performed measurement calculated with the standard speech to obtain five kinds of measure.After the fusion of five measures, random forest model was used to assessed the quality level.Finally, a neural network model based on residual structure was designed combined multimodal fusion technique, which fused the graph domain and measure domain features of the interfered speech data and performed quality level assessment.Experimental results show that the accuracy of two methods have reached more than 90%.Among them, the multimodal assessment method improves the accuracy by about 3.269% compared with the existing research methods, which proves that it has a better performance.

Cyber attacks and privacy protection distributed consensus algorithm for multi-agent systems

Ming XU, Baojun ZHANG, Yiming WU, Chenduo YING, Ning ZHENG

2023, 44(3):  117-127.  doi:10.11959/j.issn.1000-436x.2023049

Asbtract ( 333 )   HTML ( 54)   PDF (1427KB) ( 818 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To solve the consensus problem of multi-agent systems in the environment of cyber attacks and information eavesdropping, an average consensus algorithm that can effectively resist network denial of service (DoS) attacks and protect nodes’ state privacy was proposed.Firstly, combined with the characteristics of DoS attacks in networked control systems, a periodic DoS attack model related to time was constructed.Then a distributed network node information processing mechanism based on differential privacy was proposed by using the information interaction between neighbor nodes, and then it was introduced into the average consensus algorithm.Furthermore, combined with the event-triggered mechanism, a consensus algorithm method suitable for undirected communication networks under DoS attacks was proposed, and its convergence and privacy protection performance were analyzed strictly mathematically.Finally, the effectiveness of the proposed algorithm was verified by a numerical simulation experiment and hardware experiment.

Reconstruction of LDPC code check matrix based on random extraction at high bit error rate

Zhongyong WANG, Zhenghao LI, Kexian GONG, Peng SUN, Qingtao LI

2023, 44(3):  128-137.  doi:10.11959/j.issn.1000-436x.2023062

Asbtract ( 149 )   HTML ( 19)   PDF (1761KB) ( 413 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to improve the performance of the sparse check matrix reconstruction algorithm of LDPC codes at high BER, an open set recognition algorithm of the check matrix with strong fault tolerance under the condition of sufficient and insufficient number of received code words was proposed.Firstly, a new code word space was constructed by randomly extracting part bits of the code words for many times.Gaussian elimination method was used to solve the dual vector and restore the check vector in a lower dimension.Secondly, using the check vector, the proportion of error-free code groups in the received data was continuously increased by using the methods of “eliminating error code words” or“flipping the lowest unreliable bits” for iterative processing.Simulation results show that the proposed algorithm is superior to comparison algorithm under different bit error rates, different code lengths, different code rates and different number of code words.For (648,324) LDPC codes in IEEE 802.11n protocol, when the number of received code words is sufficient, the reconstruction rate of check matrix can reach more than 95% under the condition of bit error rate of 0.003.When the number of received code words is insufficient (the number of code words is 450), the reconstruction rate of check matrix can reach more than 90% under the condition of bit error rate of 0.0015.

Construction of new even-length type-II Z-complementary pair

Xiaoyu CHEN, Lianfeng SUN, Jiekun ZHANG, Yubo LI

2023, 44(3):  138-144.  doi:10.11959/j.issn.1000-436x.2023052

Asbtract ( 133 )   HTML ( 14)   PDF (746KB) ( 407 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To enrich the parameter forms of Z-complementary pair (ZCP), the constructions of new Type-II ZCP were investigated.First, Type-II $\left(2^{k+2}N,(2^{k+2}-1)N\right)-\text{ZCP}$ was proposed by using iterative method.When N= 1, Z-optimal type-II $\left(2^{k+2},(2^{k+2}-1)\right)-\text{ZCP}$ could be obtained.Then, type-II $\left(2^{k+1}N+2^{k+1},2^{k+1}N+2^{k+1}-\frac{N}{2}\right)-\text{ZCP}$ was presented by combining the iterative method with the insertion method.When N= 2, Z-optimal type-II $\left(3\times 2^{k+1},3\times 2^{k+1}-1\right)-\text{ZCP}$ could be obtained.The proposed constructions can generate new even-length ZCP, and expand the number of existing ZCP.

Topology prediction for opportunistic network based on spatiotemporal convolution

Jian SHU, Jiawei SHI, Linlan LIU, Al-Kali Manar

2023, 44(3):  145-156.  doi:10.11959/j.issn.1000-436x.2023060

Asbtract ( 179 )   HTML ( 29)   PDF (4307KB) ( 189 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The high dynamics of opportunistic network topology leads to the challenges of topology prediction.The existing research mainly focuses on the long-term spatiotemporal dependence of networks, ignoring the short-term spatiotemporal features.A topology prediction method for opportunistic network based on dynamic time warping algorithm and spatiotemporal convolution (DTW-STC) was proposed, which integrated long-short term spatiotemporal dependence in opportunistic network.The time slot was determined based on dynamic time warping algorithm, so that the opportunistic network was sliced into snapshots which topology was presented by link state matrix.Temporal convolution was employed to extract short-term temporal features.The spatiotemporal graph, representing the short-term spatiotemporal relationship, was constructed by temporal features and network changes.The short-term spatiotemporal features were captured by graph convolution.After stacks of spatiotemporal convolution, the long-short term spatiotemporal features of network were achieved.Based on the autoencoder structure, vector space transformation was realized, so that the future network topology was predicted.The results on three real opportunistic network datasets, ITC, MIT, and Asturias-er, show that the proposed DTW-STC has better prediction performance than ones of other baseline methods.

Conflict evidence combination method based on clustering weighting

Yu DONG, Youpeng ZHANG

2023, 44(3):  157-163.  doi:10.11959/j.issn.1000-436x.2023029

Asbtract ( 106 )   HTML ( 15)   PDF (801KB) ( 368 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Aiming at the failure of the combination of highly conflicting evidence, the combination of the relationship between evidences and the characteristics of evidence itself was considered, and a method of conflicting evidence combination based on clustering weighting was proposed.Firstly, the similarity coefficient was constructed to cluster the evidence, and the inter-class weight was determined by class capacity to reflect the differences among classes.Then the information entropy was used to obtain the intra-class weight of each evidence.The modified weight of the evidence was obtained by combining the inter-class weight and intra-class weight.Finally, the conflict evidence was modified according to the weight, and the combination rule of evidence theory was used to achieve the final synthesis of the modified evidence.The results of the example show that the proposed method can reasonably carry out evidence correction, and the combined results have better focusing effect than other methods.

Distributed audit causal consistency model based on biased stability

Junfeng TIAN, Qianyu YANG, Xiao Jitian

2023, 44(3):  164-177.  doi:10.11959/j.issn.1000-436x.2023059

Asbtract ( 133 )   HTML ( 15)   PDF (3787KB) ( 354 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In the distributed storage, causal consistency is favored due to the best trade-off between ease of programming and performance.To address the problem of vector-dependent tracking loss of throughput in existing causal consistency results, a distributed audit causal consistency model based on biased stability was proposed.Combined vector timestamps were used instead of full vector timestamps in query operations to reduce system management and communication overhead.Meanwhile, the causal auditing was introduced with the help of distributed associative arrays, and data dependency was refined by partitioned cooperative auditing to reduce the number of false dependency entries.Theoretical analysis and simulation results show that proposed model improves throughput by 48.26% and reduces update response latency by 16.25%.

Adaptor signature scheme based on ISRSAC digital signature algorithm

Yanshuo ZHANG, Ning LIU, Yuqi YUAN, Yatao YANG

2023, 44(3):  178-185.  doi:10.11959/j.issn.1000-436x.2023056

Asbtract ( 217 )   HTML ( 34)   PDF (732KB) ( 344 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The adapter signature scheme can provide good properties of atomic exchange in the block chain, and has been widely used in practice.A new adapter signature scheme was constructed based on a digital signature algorithm for improved security of RSA cryptosystem (ISRSAC).After proving that the proposed scheme satisfied the correctness of the pre-signature, unforgeability, adaptability of the pre-signature, extractability of the evidence and the security of the signature scheme, it was compared with the adapter signature scheme based on the SM2 digital signature algorithm, the Schnorr-based adapter signature scheme and the ECDSA-based adapter signature scheme.And these schemes were analyzed in terms of time overhead and main calculation amount.The analysis results show that the proposed scheme is similar to other schemes in terms of time overhead, but the proposed scheme expands the application scenarios of ISRSAC algorithm in the current environment, and through the combination of ISRSAC algorithm and adapter signature technology, the selection of adapter signature is more flexible and can have a wider range of applications.

Research on efficient and secure cross-domain data sharing of IoV under blockchain architecture

Xuejiao LIU, Tiancong CAO, Yingjie XIA

2023, 44(3):  186-197.  doi:10.11959/j.issn.1000-436x.2023031

Asbtract ( 517 )   HTML ( 116)   PDF (2310KB) ( 567 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

To solve the problems of data disclosure, uncontrolled data access, and inefficiency of cross-domain data sharing in the Internet of vehicles (IoV), an efficient and secure cross-domain data sharing scheme of IoV under blockchain architecture was proposed.A blockchain was maintained by trusted authorities of different trust domains.A modified ciphertext-policy attribute-based encryption scheme was adopted to encrypt data, and encrypted data was stored in interplanetary file system (IPFS) with relevant information recorded on the blockchain, constructing a fine-grained and secure cross-domain data sharing scheme based on blockchain.A verification algorithm for cross-domain access based on the garbled Bloom filter was designed, and a smart contract executed fast decryption tests based on access policies on the blockchain, improving the access efficiency of a mass of ciphertext.A cross-domain data access method based on outsourcing decryption was designed, and the trusted authorities transformed ciphertexts while performing outsourcing decryption with complex bilinear pairing calculations, reducing the time overhead of vehicle decryption.Experiment results show that the proposed scheme is superior to other schemes in the process of cross-domain ciphertext transformation and vehicle decryption, and the cross-domain data access efficiency is increased by 60% on average.

Correspondences

Deep image semantic communication model for 6G

Feibo JIANG, Yubo PENG, Li DONG

2023, 44(3):  198-208.  doi:10.11959/j.issn.1000-436x.2023050

Asbtract ( 486 )   HTML ( 116)   PDF (2900KB) ( 883 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Current semantic communication models still have some parts that can be improved in processing image data, including effective image semantic codec, efficient semantic model training, and accurate image semantic evaluation.Hence, a deep image semantic communication (DeepISC) model was proposed.The vision transformer-based autoencoder (ViTA) network was used to achieve high-quality image semantic encoding and decoding.Then, an autoencoder realized channel codec to ensure the transmission of semantics on the channel.Furthermore, the discriminator network (DSN) and ViTA’s dual network architecture were used to jointly train, thus improving the semantic accuracy of the reconstructed image.Finally, for different downstream vision tasks, different evaluation indicators of image semantics were presented.Simulation results show that compared with other schemes, DeepISC can more effectively restore the semantic features of the transmitted image, so that the reconstructed image can show the same or similar semantic results as the original image in various downstream tasks.

Robust feature selection method via joint low-rank reconstruction and projection reconstruction

Shuangyan YI, Yongsheng LIANG, Jingjing LU, Wei LIU, Tao HU, Zhenyu HE

2023, 44(3):  209-219.  doi:10.11959/j.issn.1000-436x.2023061

Asbtract ( 142 )   HTML ( 23)   PDF (3585KB) ( 407 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Aiming at the problem that current feature selection methods were still affected by noise and cannot effectively unify clustering and reconstruction effects, a robust feature selection method was proposed.A robust reconstruction error term was built by making the difference between low-rank reconstruction and projection reconstruction.After that, the features for clustering were selected from the reconstructed data instead of the original data.The learning of clean data and feature selection technique are allowed for joint learning and promote each other, thereby improving the robustness of the method on noisy data, and effectively unifying reconstruction and clustering.Compared with several kinds of graph embedding feature selection and reconstruction feature selection methods on five datasets, the experimental results showed that, except for the LUNG noise dataset, the proposed method outperforms the comparative feature selection method under both evaluation indicators (ACC and NMI).

Security analysis of public-key cryptosystems based on matrix action problem against quantum attack

Huawei HUANG

2023, 44(3):  220-226.  doi:10.11959/j.issn.1000-436x.2023064

Asbtract ( 159 )   HTML ( 27)   PDF (771KB) ( 394 )   Knowledge map   

References | Related Articles | Metrics

As a generalization of the discrete logarithm problem, semigroup action problem has important applications in the design of public-key cryptography.Public-key cryptosystems based on action problem of integer matrix semigroups on the direct product of commutative groups were analyzed.The matrix was regarded as the exponent of direct product elements, and this class of matrix action had the exponential rules similar to group.It was proved that if the matrix action was injective or the number of generators of the hidden subgroup was less than or equal to the square of the order of the matrix, the matrix action problem could be reduced in polynomial time to the hidden subgroup problem of the direct sum of the additive group of the matrices.And it was proved that commutative matrix action problem could also be reduced to hidden subgroup problem of the direct sum of the additive group of the matrices in polynomial time.The cryptosystems based on this class of matrix action problem cannot against quantum attacks.This conclusion has theoretical significance in the design of public-key cryptography against quantum attacks.