Table of Content

10 November 2023, Volume 43 Issue 11 Catalog Download

Previous Issue    Next Issue

National Open Distributed and Parallel Computing Conference 2022 (DPCS 2022）

Current research status and challenges of blockchain in supply chain applications

Lina GE, Jingya XU, Zhe WANG, Guifen ZHANG, Liang YAN, Zheng HU

2023, 43(11):  3315-3326.  DOI: 10.11772/j.issn.1001-9081.2022111758

Asbtract ( )   PDF (2371KB) ( )  

References | Related Articles | Metrics

The supply chain faces many challenges in the development process， including how to ensure the authenticity and reliability of information as well as the security of the traceability system in the process of product traceability， the security of products in the process of logistics， and the trust management in the financing process of small and medium enterprises. With characteristics of decentralization， immutability and traceability， blockchain provides efficient solutions to supply chain management， but there are some technical challenges in the actual implementation process. To study the applications of blockchain technology in the supply chain， some typical applications were discussed and analyzed. Firstly， the concept of supply chain and the current challenges were briefly introduced. Secondly， problems faced by blockchain in three different supply chain fields of information flow， logistics flow and capital flow were described， and a comparative analysis of related solutions was given. Finally， the technical challenges faced by blockchain in the practical applications of supply chain were summarized， and future applications were prospected.

Parallel computing algorithm of grid-based distributed Xin’anjiang hydrological model

Qian LIU, Yangming ZHANG, Dingsheng WAN

2023, 43(11):  3327-3333.  DOI: 10.11772/j.issn.1001-9081.2022111760

Asbtract ( )   HTML ( )   PDF (2494KB) ( )  

Figures and Tables | References | Related Articles | Metrics

In recent years， the Grid-based distributed Xin’anjiang hydrological Model （GXM） has played an important role in flood forecasting， but when simulating the flooding process， due to the vast amount of data and calculation of the model， the computing time of GXM increases exponentially with the increase of the model warm-up period， which seriously affects the computational efficiency of GXM. Therefore， a parallel computing algorithm of GXM based on grid flow direction division and dynamic priority Directed Acyclic Graph （DAG） scheduling was proposed. Firstly， the model parameters， model components， and model calculation process were analyzed. Secondly， a parallel algorithm of GXM based on grid flow direction division was proposed from the perspective of spatial parallelism to improve the computational efficiency of the model. Finally， a DAG task scheduling algorithm based on dynamic priority was proposed to reduce the occurrence of data skew in model calculation by constructing the DAG of grid computing nodes and dynamically updating the priorities of computing nodes to achieve task scheduling during GXM computation. Experimental results on Dali River basin of Shaanxi Province and Tunxi basin of Anhui Province show that compared with the traditional serial computing method， the maximum speedup ratio of the proposed algorithm reaches 4.03 and 4.11， respectively， the computing speed and resource utilization of GXM were effectively improved when the warm-up period is 30 days and the data resolution is 1 km.

Modeling and simulation of CPS based on object spatiotemporal Petri net

Liangliang DENG, Lichen ZHANG, Wenchao JIANG

2023, 43(11):  3334-3339.  DOI: 10.11772/j.issn.1001-9081.2022111759

Asbtract ( )   HTML ( )   PDF (2582KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Cyber-Physical System （CPS） is a distributed real-time feedback system that integrates computing， control， communication and physical elements， but the traditional modeling methods cannot meet the high requirements of CPS for spatiotemporal performance. To address this problem， a modeling method for Duration-Space Object Petri Net （DS-OPN） was proposed. Firstly， the object-oriented encapsulation technology and spatiotemporal elements were integrated into the Petri net， and the spatial and temporal description rules were designed to encapsulate scenario elements under the same object into the same object subnet system model. Secondly， the aggregation rules were defined to aggregate various subnet models to enable these models to describe the object change process in the CPS physical topology environment. Finally， taking the traffic CPS as an example， the dynamic behavior of the autonomous control overtaking system was modeled and simulated； at the same time， the coverability tree and the incidence matrix of the model were established to verify the model’s reachability and safety. Experimental results show that the model modeled by the proposed method has a clear representation of the logical structure of the system flow as well as accurate calculation of spatiotemporal factors， and meets the requirements of CPS in terms of real-time performance and security， which verifies the effectiveness and security of the modeling method.

Recommendation rating prediction algorithm based on user interest concept lattice reduction

Xuejian ZHAO, Hao LI, Haotian TANG

2023, 43(11):  3340-3345.  DOI: 10.11772/j.issn.1001-9081.2022121839

Asbtract ( )   HTML ( )   PDF (1411KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The performance of the recommendation systems is restricted by data sparsity， and the accuracy of prediction can be effectively improved by reasonably filling the missing values in the rating matrix. Therefore， a new algorithm named Recommendation Rating Prediction based on Concept Lattice Reduction （RRP-CLR） was proposed. RRP-CLR algorithm was composed of nearest neighbor selection module and rating prediction module， which were respectively responsible for generating reduced nearest neighbor set and realizing rating prediction and recommendation. In the nearest neighbor selection module， the user rating matrix was transformed into a binary matrix， which was regarded as the user interest formal background. Then the formal background reduction rules and concept lattice redundancy concept deletion rules were proposed to improve the efficiency of generating reduced nearest neighbors. In the rating prediction module， a new user similarity calculation method was proposed to eliminate the impact of rating deviations caused by user’s subjective factors on similarity calculation. When the number of common rating items of two users was less than a specific threshold， the similarity was scaled appropriately to make the similarity between users more consistent with the real situation. Experimental results show that compared with PC?UCF （User-based Collaborative Filtering recommendation algorithm based on Pearson Coefficient） and RRP-UICL （Recommendation Rating Prediction method based on User Interest Concept Lattice）， RRP-CLR algorithm has smaller Mean Absolute Error （MAE） and Root Mean Squared Error （RMSE）， and better rating prediction accuracy and stability.

Advertising recommendation algorithm based on differential privacy

Lei TIAN, Lina GE

2023, 43(11):  3346-3350.  DOI: 10.11772/j.issn.1001-9081.2023010106

Asbtract ( )   HTML ( )   PDF (1100KB) ( )  

Figures and Tables | References | Related Articles | Metrics

With the rapid development of the mobile Internet industry， user data and browsing data have increased significantly， so it is extremely important to accurately grasp the potential needs of users and improve the effect of advertisement recommendation. As a relatively advanced recommendation method at present， DeepFM model can extract various complexity features from the original features， but the model does not protect the data. In order to realize the privacy protection in DeepFM model， a new DeepFM model based on Differential Privacy （DP） was proposed， namely DP-DeepFM. The Gaussian noise was added to Adam optimization algorithm in the training process of DP-DeepFM and the gradient clipping was performed to prevent the addition of excessive noise causing poor model performance. Experimental results on advertising dataset Criteo show that compared with DeepFM， DP-DeepFM only has the accuracy decreased by 0.44 percentage points， but it provides differential privacy protection and is more secure.

General multi-unit false-name-proof auction mechanism for cloud computing

Kun YOU, Qinhui WANG, Xin LI

2023, 43(11):  3351-3357.  DOI: 10.11772/j.issn.1001-9081.2022111705

Asbtract ( )   HTML ( )   PDF (1731KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Aiming at the problem of resource auction mechanism in cloud environment， a more General multi-unit FAlse-name-proof auction mechanism for vIrTual macHine allocation （GFAITH） was studied and designed. First， the system model was formally defined. Then， around the design goals of being truthfulness and false-name-proof， it was proved that when considering the diversity of user demands， a new form of cheating， Demand-Reduction （DR） cheating， would emerge， which could destroy the truthful and false-name-proof properties， and the experimental results show that it would seriously affect the system performance. Based on the above， the GFAITH was proposed to achieve the design goals in three stages： user pre-processing， pre-allocation and pricing， and resisting demand reduction cheating. It is theoretical proved that the resource allocation of GFAITH is feasible and able to resist false-name-proof. Experimental results show that GFAITH can effectively guarantee the performance of the system from indicators such as revenue and social wealth， verifying the effectiveness and efficiency of the proposed mechanism.

Certificateless conditional privacy-preserving authentication scheme for VANET

Guishuang XU, Xinchun YIN

2023, 43(11):  3358-3367.  DOI: 10.11772/j.issn.1001-9081.2022111757

Asbtract ( )   HTML ( )   PDF (867KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Vehicular Ad-hoc NETwork （VANET） is vital for constructiong intelligent transportation systems because of obvious advantages in sharing traffic data， improving driving efficiency and reducing traffic accidents. Meanwhile， problems such as secure communication of vehicle-to-vehicle and vehicle-to-infrastructure， privacy-preserving of vehicles （e.g.， identity privacy， location privacy）， and efficient authentication of traffic messages need to be solved urgently. To achieve a trade-off between security and efficiency， firstly， the recently proposed scheme， namely Conditional Privacy-Preserving CertificateLess Aggregate Signature scheme （CPP-CLAS）， was analyzed and proved to be unable to resist the public key replacement attack. Then， based on this scheme， a new certificateless conditional privacy-preserving authentication scheme for VANET was proposed， in which the secure channels were not required during partial private key generation of vehicles. In addition， aggregate verification and batch verification were employed to verify a batch of signatures in the scheme. Finally， the proposed scheme was proved to have unforgeability under random oracle model. Performance analysis show that compared with the similar schemes， the proposed scheme improves the computational efficiency of the signature phase by at least 66.76% and reduces the communication bandwidth demand by at least 16.67% without increasing the verification overhead， verifying that the proposed scheme is more suitable for resource-constrained VANET.

Efficient certificateless ring signature scheme based on elliptic curve

Xiuping ZHU, Yali LIU, Changlu LIN, Tao LI, Yongquan DONG

2023, 43(11):  3368-3374.  DOI: 10.11772/j.issn.1001-9081.2022111801

Asbtract ( )   HTML ( )   PDF (740KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Ring signature is widely used to solve the problems of user identity and data privacy disclosure because of its spontaneity and anonymity； and certificateless public key cryptosystem can not only solve the problem of key escrow， but also do not need the management of public key certificates； certificateless ring signature combines the advantages of both of the above mentioned， and has extensive research significance， but most of the existing certificateless ring signature schemes are based on the calculation of bilinear pairings and modular exponentiation， which are computationally expensive and inefficient. In order to improve the efficiency of signature and verification stages， a new Efficient CertificateLess Ring Signature （ECL-RS） scheme was proposed， which used elliptic curve with low computational cost， high security and good flexibility. The security statute of ECL-RS scheme stems from a discrete logarithm problem and a Diffie-Hellman problem， and the scheme is proved to be resistant to public key substitution attacks and malicious key generation center attacks under Random Oracle Model （ROM） with unforgeability and anonymity. Performance analysis shows that ECL-RS scheme only needs （n+2） （n is the number of ring members） elliptic curve scalar multiplication and scalar addition operations as well as （n+3） one-way hash operations， which has lower computational cost and higher efficiency while ensuring security.

Vehicle RKE two-factor authentication protocol resistant to physical cloning attack

Changgeng LIU, Yali LIU, Qipeng LU, Tao LI, Changlu LIN, Yi ZHU

2023, 43(11):  3375-3384.  DOI: 10.11772/j.issn.1001-9081.2022111802

Asbtract ( )   HTML ( )   PDF (1299KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Attackers can illegally open a vehicle by forgeing the Radio Frequency IDentification （RFID） signal sent by the vehicle remote key. Besides， when the vehicle remote key is lost or stolen， the attacker can obtain the secret data inside the vehicle remote key and clone a usable vehicle remote key， which will threaten the property and privacy security of the vehicle owner. Aiming at the above problems， a Vehicle RKE Two-Factor Authentication （VRTFA） protocol for vehicle Remote Keyless Entry （RKE） that resists physical cloning attack was proposed. The protocol is based on Physical Uncloneable Function （PUF） and biological fingerprint feature extraction and recovery functions， so that the specific hardware physical structure of the legal vehicle remote key cannot be forged. At the same time， the biological fingerprint factor was introduced to build a two-factor authentication protocol， thereby solving the security risk of vehicle remote key theft， and further guaranteeing the secure mutual authentication of vehicle RKE system. Security analysis results of the protocol using BAN logic show that VRTFA protocol can resist malicious attacks such as forgery attack， desynchronization attack， replay attack， man-in-the-middle attack， physical cloning attack， and full key leakage attack， and satisfy the security attributes such as forward security， mutual authentication， data integrity， and untraceability. Performance analysis results show that VRTFA protocol has stronger security and privacy and better practicality than the existing RFID authentication protocols.

Artificial intelligence

Review of multi-modal medical image segmentation based on deep learning

Meng DOU, Zhebin CHEN, Xin WANG, Jitao ZHOU, Yu YAO

2023, 43(11):  3385-3395.  DOI: 10.11772/j.issn.1001-9081.2022101636

Asbtract ( )   HTML ( )   PDF (3904KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Multi-modal medical images can provide clinicians with rich information of target areas （such as tumors， organs or tissues）. However， effective fusion and segmentation of multi-modal images is still a challenging problem due to the independence and complementarity of multi-modal images. Traditional image fusion methods have difficulty in addressing this problem， leading to widespread research on deep learning-based multi-modal medical image segmentation algorithms. The multi-modal medical image segmentation task based on deep learning was reviewed in terms of principles， techniques， problems， and prospects. Firstly， the general theory of deep learning and multi-modal medical image segmentation was introduced， including the basic principles and development processes of deep learning and Convolutional Neural Network （CNN）， as well as the importance of the multi-modal medical image segmentation task. Secondly， the key concepts of multi-modal medical image segmentation was described， including data dimension， preprocessing， data enhancement， loss function， and post-processing， etc. Thirdly， different multi-modal segmentation networks based on different fusion strategies were summarized and analyzed. Finally， several common problems in medical image segmentation were discussed， the summary and prospects for future research were given.

Point cloud classification and segmentation based on Siamese adaptive graph convolution algorithm

Weigang LI, Ting CHEN, Zhiqiang TIAN

2023, 43(11):  3396-3402.  DOI: 10.11772/j.issn.1001-9081.2022101552

Asbtract ( )   HTML ( )   PDF (2328KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Point cloud data has sparsity， irregularity， and permutation invariance， and lacks topological information， which makes it difficult to extract features of point cloud. Therefore， a Siamese Adaptive Graph Convolution Algorithm （SAGCA） was proposed for point cloud classification and segmentation. Firstly， the topological relationships between irregular and sparse point cloud features were mined by constructing feature relationship graph. Then， the Siamese composition idea of sharing convolution learning weights was introduced to ensure the permutation invariance of point cloud data and make the topological relationship expression more accurate. Finally， SAGCA was combined with various deep learning networks for processing point cloud data by both global and local combination methods， thereby enhancing the feature extraction ability of the network. Comparison results with PointNet++ benchmark network of the classification， object part segmentation and scene semantic segmentation experiments on ScanObjectNN， ShapeNetPart and S3DIS datasets， respectively， show that， based on the same dataset and evaluation criteria， SAGCA has the class mean Accuracy （mAcc） of classification increased by 2.80 percentage points， the overall class average Intersection over Union （IoU） of part segmentation increased by 2.31 percentage points， and the class mean Intersection over Union （mIoU） of scene semantic segmentation increased by 2.40 percentage points， verifying that SAGCA can effectively enhance the feature extraction ability of the network and is suitable for multiple point cloud classification and segmentation tasks.

Human pose transfer model combining convolution and multi-head attention

Hong YANG, He ZHANG, Shaoning JIN

2023, 43(11):  3403-3410.  DOI: 10.11772/j.issn.1001-9081.2022111707

Asbtract ( )   HTML ( )   PDF (2734KB) ( )  

Figures and Tables | References | Related Articles | Metrics

For a given reference image of a person， the goal of Human Pose Transfer （HPT） is to generate an image of that person in any arbitrary pose. Many existing related methods fail to capture the details of a person’s appearance and have difficulties in predicting invisible regions， especially for complex pose transformation， and it is difficult to generate a clear and realistic person’s appearance. To address the above problems， a new HPT model that integrated convolution and multi-head attention was proposed. Firstly， the Convolution-Multi-Head Attention （Conv-MHA） block was constructed by fusing the convolution and multi-head attention， then it was used to extract rich contextual features. Secondly， to improve the learning ability of the proposed model， the HPT network was constructed by using Conv-MHA block. Finally， the self-reconstruction of the reference image was introduced as an auxiliary task to make the model more fully utilized its performance. The Conv-MHA-based human pose transfer model was validated on DeepFashion and Market-1501 datasets， and the results on DeepFashion test dataset show that it outperforms the state-of-the-art human pose transfer model， DPTN （Dual-task Pose Transformer Network）， in terms of Structural SIMilarity （SSIM）， Learned Perceptual Image Patch Similarity （LPIPS） and FID （Fréchet Inception Distance） indicators. Experimental results show that the Conv-MHA module， which integrates convolution and multi-head attention mechanism， can improve the representation ability of the model， capture the details of person’s appearance more effectively， and improve the accuracy of person image generation.

Shared transformation matrix capsule network for complex image classification

Kai WEN, Xiao XUE, Juan JI

2023, 43(11):  3411-3417.  DOI: 10.11772/j.issn.1001-9081.2022101596

Asbtract ( )   HTML ( )   PDF (2309KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Concerning the problems of poor classification performance and high computational overhead of Capsule Network （CapsNet） on complex images with background noise information， an improved capsule network model based on attention mechanism and weight sharing was proposed， called Shared Transformation Matrix CapsNet （STM-CapsNet）. The proposed model mainly includes the following improvement. 1） An attention module was introduced into the feature extraction layer of CapsNet， which enabled low-level capsules to focus on entity features related to the classification task. 2） Low-level capsules with close spatial positions were divided into several groups， and each group of low-level capsules was mapped to high-level capsules by sharing transformation matrices， which reduced computational overhead and improved model robustness. 3） The L₂ regularization term was added to margin loss and reconstruction loss to prevent model overfitting. Experimental results on three complex image datasets including CIFAR10， SVHN （Street View House Number） and FashionMNIST show that， the above improvements are effective in enhacing the model performance； when the number of iterations is 3， and the number of shared transformation matrices is 5， the average accuracies of STM-CapsNet are 85.26%， 93.17% and 94.96% respectively， the average parameter amount is 8.29 MB， verifying that STM-CapsNet has better performance compared with the baseline models.

EEG classification based on channel selection and multi-dimensional feature fusion

Shuying YANG, Haiming GUO, Xin LI

2023, 43(11):  3418-3427.  DOI: 10.11772/j.issn.1001-9081.2022101590

Asbtract ( )   HTML ( )   PDF (3363KB) ( )  

Figures and Tables | References | Related Articles | Metrics

To solve the problems of the mutual interference of multi-channel ElectroEncephaloGraphy （EEG）， the different classification results caused by individual differences， and the low recognition rate of single domain features， a method of channel selection and feature fusion was proposed. Firstly， the acquired EEG was preprocessed， and the important channels were selected by using Gradient Boosting Decision Tree （GBDT）. Secondly， the Generalized Predictive Control （GPC） model was used to construct the prediction signals of important channels and distinguish the subtle differences among multi-dimensional correlation signals， then the SE-TCNTA （Squeeze and Excitation block-Temporal Convolutional Network-Temporal Attention） model was used to extract temporal features between different frames. Thirdly， the Pearson correlation coefficient was used to calculate the relationship between channels， the frequency domain features of EEG and the control values of prediction signals were extracted as inputs， the spatial graph structure was established， and the Graph Convolutional Network （GCN） was used to extract the features of frequency domain and spatial domain. Finally， the above two features were input to the fully connected layer for feature fusion in order to realize the classification of EEG. Experimental results on public dataset BCICIV_2a show that in the case of channel selection， compared with the first EEG-inception model for ERP detection and DSCNN （Shallow Double-branch Convolutional Neural Network） model that also uses double branch feature extraction， the proposed method has the classification accuracy increased by 1.47% and 1.69% respectively， and has the Kappa value increased by 1.25% and 2.53% respectively. The proposed method can improve the classification accuracy of EEG and reduce the influence of redundant data on feature extraction， so it is more suitable for Brain-Computer Interface （BCI） systems.

Cross-model universal perturbation generation method based on geometric relationship

Jici ZHANG, Chunlong FAN, Cailong LI, Xuedong ZHENG

2023, 43(11):  3428-3435.  DOI: 10.11772/j.issn.1001-9081.2022111677

Asbtract ( )   HTML ( )   PDF (3981KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Adversarial attacks add designed perturbations to the input samples of neural network models to make them output wrong results with high confidence. The research on adversarial attacks mainly aim at the application scenarios of a single model， and the attacks on multiple models are mainly realized through cross-model transfer attacks， but there are few studies on universal cross-model attack methods. By analyzing the geometric relationship of multi-model attack perturbations， the orthogonality of the adversarial directions of different models and the orthogonality of the adversarial direction and the decision boundary of a single model were clarified， and the universal cross-model attack algorithm and corresponding optimization strategy were designed accordingly. On CIFAR10， SVHN datasets and six common neural network models， the proposed algorithm was verified by multi-angle cross-model adversarial attacks. Experimental results show that the attack success rate of the algorithm in a given experimental scenario is 1.0， and the L2-norm is not greater than 0.9. Compared with the cross-model transfer attack， the proposed algorithm has the average attack success rate on the six models increased by up to 57% and has better universality.

Universal perturbation generation method of neural network based on differential evolution

Qianshun GAO, Chunlong FAN, Yanda LI, Yiping TENG

2023, 43(11):  3436-3442.  DOI: 10.11772/j.issn.1001-9081.2022111733

Asbtract ( )   HTML ( )   PDF (1601KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Aiming at the problem that the universal perturbation search in HGAA （Hyperspherical General Adversarial Attacks） algorithm is always limited to the spatial spherical surface， and it does not have the ability to search the space inside the sphere， a differential evolution algorithm based on hypersphere was proposed. In the algorithm， the search space was expanded to the interior of the sphere， and Differential Evolution （DE） algorithm was used to search the optimal sphere， so as to generate universal perturbations with higher fooling rate and lower modulus length on this sphere. Besides， the influence of key parameters such as the number of populations on the algorithm was analyzed， and the performance of the universal perturbations generated by the algorithm on different neural network models was tested. The algorithm was verified on CIFAR10 and SVHN image classification datasets， and the fooling rate of the algorithm was increased by up to 11.8 percentage points compared with that of HGAA algorithm. Experimental results show that this algorithm extends the universal perturbation search space of the HGAA algorithm， reduces the modulus length of universal perturbation， and improves the fooling rate of universal perturbations.

Deep review attention neural network model for enhancing explainability of recommendation system

Chuyuan WEI, Mengke WANG, Chuanhao HU, Guangqi ZHANG

2023, 43(11):  3443-3448.  DOI: 10.11772/j.issn.1001-9081.2022101628

Asbtract ( )   HTML ( )   PDF (1652KB) ( )  

Figures and Tables | References | Related Articles | Metrics

In order to improve the explainability of Recommendation System （RS）， break the inherent limitations of recommendation system and enhance the user’s trust and satisfaction on recommender systems， a Deep Review Attention Neural Network （DRANN） model with enhanced explainability was proposed. Based on the potential relationships between users and items on text reviews， the rich semantic information in user reviews and item reviews was used to predict users’ interest preferences and sentiment tendencies by the proposed model. Firstly， a Text Convolutional Neural Network （TextCNN） was used to do shallow feature extraction for word vectors. Then， the attention mechanism was used to assign weights to comment data and filter invalid comment information. At the same time， the deep autoencoder module was constructed to reduce the dimension of high-dimensional sparse data， remove interference information， learn deep semantic representation， and enhance the explainability of recommendation model. Finally， the prediction score was obtained through the prediction layer. Experimental results on the four public data sets including Patio， Automotive， Musical Instrument （M?I） and Beauty show that DRANN model has the smallest Root Mean Square Error （RMSE） compared with Probabilistic Matrix Factorization （PMF）， Single Value Decomposition++ （SVD++）， Deep Cooperative Neural Network （DeepCoNN）， Tree-enhanced Embedding Model （TEM）， DeepCF （Deep Collaborative Filtering） and DER（Dynamic Explainable Recommender）， verifying its effectiveness in improving performance and the feasibility of the adopted explanation strategy.

Data science and technology

Multiple clustering algorithm based on dynamic weighted tensor distance

Zhuangzhuang XUE, Peng LI, Weibei FAN, Hongjun ZHANG, Fanshuo MENG

2023, 43(11):  3449-3456.  DOI: 10.11772/j.issn.1001-9081.2022101626

Asbtract ( )   HTML ( )   PDF (2437KB) ( )  

Figures and Tables | References | Related Articles | Metrics

When measuring the importance of attributes in Tensor-based Multiple Clustering algorithm （TMC）， the relevance of attribute combinations within object tensors are ignored， and the selected and unselected feature space are incompletely separated because of the fixed weight strategy under different feature space selection. For above problems， a Multiple Clustering algorithm based on Dynamic Weighted Tensor Distance （DWTD-MC） was proposed. Firstly， a self-association tensor model was constructed to improve the accuracy of attribute importance measurement of each feature space. Then， a multi-view weight tensor model was built to meet the task requirements of multiple clustering analysis by dynamic weighting strategy under different feature space selection. Finally， the dynamic weighted tensor distance was used to measure the similarity of data points， generating multiple clustering results. Simulation results on real datasets show that DWTD-MC outperforms comparative algorithms such as TMC in terms of Jaccard Index （JI）， Dunn Index （DI）， Davies-Bouldin index （DB） and Silhouette Coefficient （SC）. It can obtain high quality clustering results while maintaining low redundancy among clustering results， as well as meeting the task requirements of multiple clustering analysis.

Nonuniform time slicing method based on prediction of community variance

Xiangyu LUO, Ke YAN, Yan LU, Tian WANG, Gang XIN

2023, 43(11):  3457-3463.  DOI: 10.11772/j.issn.1001-9081.2022111736

Asbtract ( )   HTML ( )   PDF (1001KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Time slicing methods in dynamic networks greatly influence the accuracy of community evolution analysis results. As communities vary nonlinearly with time and network topology， both the existing uniform time slicing method and network topology variance-based nonuniform time slicing method are unsatisfactory in capturing community evolution events. Therefore， a nonuniform time slicing method based on prediction of community variance was proposed， where the community variance is quantitatively described by the difference between the community modularity expected to be achieved by the updated network and the community modularity obtained by directly applying the community detection results of the network before changing. Firstly， the prediction model of community modularity was established on the basis of time series analysis. Secondly， with the established model， the expected community modularity of the updated network was predicted， and the prediction value of community variance was obtained. Finally， once the prediction value surpassed a previously set threshold， a new time slice was generated. Experimental results on two real network datasets show that compared with the traditional uniform time slicing method and the nonuniform time slicing method based on network topology variance， on the dynamic network dataset Arxiv HEP-PH， the proposed method identifies community disappearance events 1.10 days and 1.30 days earlier， respectively， and identifies the community forming events 8.34 days and 3.34 days earlier， respectively， and the total number of identified community shrinking and growing events increased by 10 and 1 respectively. On Sx?MathOverflow dataset， the proposed method identifies community disappearance events 3.30 days and 1.80 days earlier， and identifies the community forming events 6.41 days and 2.97 days earlier respectively， and the total number of identified community shrinking and growing events increased by 15 and 7， respectively.

Social recommendation by enhanced GNN with heterogeneous relationship

Yonggui WANG, Qiwen SHI

2023, 43(11):  3464-3471.  DOI: 10.11772/j.issn.1001-9081.2022111774

Asbtract ( )   HTML ( )   PDF (1897KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Social recommendation aims to use users’ social attributes to recommend potential items of interest， which effectively alleviates the problems of data sparsity and cold start. However， the existing social recommendation algorithms mainly focus on studying a single social relationship， and social attributes are difficult to fully participate in calculations， so that there are problems of failure to fully explore social heterogeneous relationships and poor quality of node feature representation. Therefore， an enhanced GNN model for social recommendation with Heterogeneous Relationship （HR-GNN） was proposed. In HR-GNN， Graph Convolutional Network （GCN） was used to aggregate user and item node information to generate query embeddings for node information query； the social heterogeneity relationships were explored by neighbor sampling strategy that combines sampling probabilities with consistency scores among neighbor nodes； and the node information was aggregated by self-attention mechanism to improve the quality of user and item feature representation. Experimental results on two real-world datasets demonstrate that in comparison with baseline algorithms， the proposed algorithm has significant improvements in both Mean Absolute Error （MAE） and Root Mean Square Error （RMSE）， and they are reduced by at least 1.80% and 1.35% on Ciao dataset and at least 2.80% and 3.18% on Epinions dataset， verifying the effectiveness of HR-GNN model.

Cyber security

Survey of semantics-based location privacy protection

Wenxuan LI, Hao WU, Changsong LI

2023, 43(11):  3472-3483.  DOI: 10.11772/j.issn.1001-9081.2022101612

Asbtract ( )   HTML ( )   PDF (2072KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The arrival of the 5G era makes Location-Based Service （LBS） more widely used， but users also have to face many privacy leakage issues caused by LBS when they enjoy the great convenience brought by LBS. In order to strengthen the security of anonymity， improve data utility， resist attacks with certain background knowledge and protect users’ sensitive information， researchers proposed the semantics-based location privacy protection mechanism. Firstly， the structure of location privacy protection systems and traditional protection technologies were introduced. Then， several typical privacy leakage and attack modes based on location semantics were analyzed， the location privacy protection requirements combined with location semantics were given， and the key technologies and achievements in the latest research of semantics-based location privacy protection from two aspects of single-point location privacy protection and trajectory privacy protection were summarized. Finally， the future technological development trend and the next research work were prospected.

Improved practical Byzantine fault tolerance algorithm based on verifiable delay function

Chundong WANG, Xin JIANG

2023, 43(11):  3484-3489.  DOI: 10.11772/j.issn.1001-9081.2022111708

Asbtract ( )   HTML ( )   PDF (2473KB) ( )  

Figures and Tables | References | Related Articles | Metrics

To solve the problems of unreasonable primary node selection and high transaction delay in Practical Byzantine Fault Tolerance （PBFT） consensus mechanism， an improved PBFT consensus mechanism based on Verifiable Delay Function （VDF） was proposed， called VPBFT. Firstly， a voting mechanism was introduced into original PBFT algorithm to select nodes， which were divided into ordinary nodes， voting nodes， backup nodes and consensus nodes according to random voting results. Secondly， the primary node selection mechanism of PBFT algorithm was improved by using VDF for primary node selection， and random numbers were generated by the hash value of the previous block and the user’s private key to increase the unpredictability of the primary node and ensure the consensus security. Finally， the consensus process of PBFT algorithm was optimized by simplifying consensus process into three stages， thereby reducing the algorithm complexity and communication overhead. Experimental results show that the proposed VPBFT outperforms the original PBFT algorithm in terms of security and consensus performance.

SM4 resistant differential power analysis lightweight threshold implementation

Jinwei PU, Qingjian GAO, Xin ZHENG, Yinghui XU

2023, 43(11):  3490-3496.  DOI: 10.11772/j.issn.1001-9081.2022101579

Asbtract ( )   HTML ( )   PDF (3082KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Aiming at the problems of large area and large consumption of fresh randomness in Threshold Implementation （TI） of SM4， an improved threshold implementation scheme of SM4 was proposed. In the case of satisfying the threshold implementation theory， the operation of S-box nonlinear inversion was shared with no fresh randomness， and a domain-oriented multiplication mask scheme was introduced to reduce the fresh randomness consumption of S-box to 12 bits. Based on the idea of the pipeline， a new SM4 serial architecture with 8-bit data width was designed. The threshold implementation of S-box was reused， and the linear function of SM4 was optimized to make the area of threshold implementation of SM4 more compact， only 6 513 GE. In comparison with the TI scheme of SM4 with 128-bit data width， the area of the proposed scheme is reduced by more than 63.7%， and there is a better trade-off between speed and area. The side-channel experimental results show that the proposed scheme has the capability of anti-first-order Differential Power Analysis （DPA）.

DDoS attack detection by random forest fused with feature selection

Jingcheng XU, Xuebin CHEN, Yanling DONG, Jia YANG

2023, 43(11):  3497-3503.  DOI: 10.11772/j.issn.1001-9081.2022111792

Asbtract ( )   HTML ( )   PDF (1450KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Exsiting machine learning-based methods for Distributed Denial-of-Service （DDoS） attack detection continue to increase in detection difficulty and cost when facing more and more complex network traffic and constantly increased data structures. To address these issues， a random forest DDoS attack detection method that integrates feature selection was proposed. In this method， the mean impurity algorithm based on Gini coefficient was used as the feature selection algorithm to reduce the dimensionality of DDoS abnormal traffic samples， thereby reducing training cost and improving training accuracy. Meanwhile， the feature selection algorithm was embedded into the single base learner of random forest， and the feature subset search range was reduced from all features to the features corresponding to a single base learner， which improved the coupling of the two algorithms and improved the model accuracy. Experimental results show that the model trained by the random forest DDoS attack detection method that integrates feature selection has a recall increased by 21.8 percentage points and an F1-score increased by 12.0 percentage points compared to the model before improvement under the premise of limiting decision tree number and training sample size， and both of them are also better than those of the traditional random forest detection scheme.

Distribution network operation exception management mechanism based on blockchain

Hongliang TIAN, Ping GE, Mingjie XIAN

2023, 43(11):  3504-3509.  DOI: 10.11772/j.issn.1001-9081.2022111665

Asbtract ( )   HTML ( )   PDF (2084KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The information interaction between operation anomalies and treatments is usually completed by operators to ensure the stable operation of distribution networks， but it is vulnerable to the subjectivity of operators， resulting in the mismatch between treatments and operation anomalies， and the lack of guarantee of information security for the interaction process. Therefore， a blockchain-based network model for distribution network exception management， Exception Management Blockchain Network （EMBN）， was proposed， as well as an improved three-line defense model for distribution network. Firstly， according to the tamper-proof and traceable characteristics of blockchain， an Anomaly Index Blockchain （AIB） was constructed， and appropriate treatments were found to solve operation anomalies based on the latest information in the block. Secondly， an Exception Interact Blockchain （EIB） was constructed to monitor the interaction process of operation anomalies and treatments， and ensure the implementation of treatments. Finally， the EMBN was applied to the three-lines of defense in traditional distribution network， and the intelligent contract was combined to realize adaptive detection and anomaly response of the distribution network. Simulation results show that， facing the complicated distribution network environment， EMBN can match treatments and operation anomalies without the influence by subjectivity of operators； compared with the traditional distribution network， EMBN has the advantage in the information security of information interaction.

Unsupervised log anomaly detection model based on CNN and Bi-LSTM

Chunyong YIN, Yangchun ZHANG

2023, 43(11):  3510-3516.  DOI: 10.11772/j.issn.1001-9081.2022111738

Asbtract ( )   HTML ( )   PDF (1759KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Logs can record the specific status of the system during the operation， and automated log anomaly detection is critical to network security. Concerning the problem of low accuracy in anomaly detection caused by the evolution of log sentences over time， an unsupervised log anomaly detection model LogCL was proposed. Firstly， the log parsing technique was used to convert semi-structured log data into structured log templates. Secondly， the sessions and fixed windows were employed to divide log events into log sequences. Thirdly， quantitative characteristics of the log sequences were extracted， natural language processing technique was used to extract semantic features of log templates， and Term Frequency-Inverse Word Frequency （TF-IWF） algorithm was utilized to generate weighted sentence embedding vectors. Finally， the feature vectors were input into a parallel model based on Convolutional Neural Network （CNN） and Bi-directional Long Short-Term Memory （Bi-LSTM） network for detection. Experimental results on two public real datasets show that the proposed model improves the anomaly detection F1-score by 3.6 and 2.3 percentage points respectively compared with the baseline model LogAnomaly. Therefore， LogCL can perform effectively on log anomaly detection.

Advanced computing

Programming model for domestic high-performance many-core processor

Hu CHEN, Pengling ZHOU

2023, 43(11):  3517-3526.  DOI: 10.11772/j.issn.1001-9081.2022101548

Asbtract ( )   HTML ( )   PDF (3529KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Programming on domestic high-performance many-core processors has requirement of using the lowest-level interface to develop software， making programming and debugging very difficult. Moreover， the limitations of programming models for high-performance software on these platforms and the absence of common computing software are identified as factors that contribute to repetitive development work. Aiming at the above problems， a generalized programming model and corresponding support library were realized： on the one hand， the thread-level parallelism of domestic high-performance many-core processors based on the message queue mechanism was developed； on the other hand， the data-level parallelism on slave cores based on the Single Instruction Multiple Data （SIMD） programming model was developed. Firstly， the architecture of the domestic high-performance multicore processor was abstracted. Then， a message queue mechanism was designed for the proposed model， along with a set of heterogeneous parallel programming interfaces， including system parameter interface， slave core thread control interface， message queue interface， and SIMD abstraction interface. Finally， a new software development model and methodology for high-performance computing were formed on the basis of the above， which was convenient for users to develop parallel computing software based on domestic high-performance many-core processors. The results of performance transmission test show that the transmission bandwidth of the proposed model on domestic many-core processors generally reaches 90% of the peak DMA （Direct Memory Access） bandwidth when a few multi-cores are turned on； and that the transmission bandwidth of the message queue model generally reaches 70% of the peak DMA bandwidth when a large number of multi-cores are turned on. In matrix multiplication experiments， the performance of the proposed model reaches 90% of the performance of the system’s original primitives for transferring matrices and calculating them； in password guessing system， the performance of the proposed model code is basically the same as that of the code developed by using the lowest-level interface directly. The proposed generalized programming model and support framework make the High Performance Computing （HPC） software development easier and more portable， which can help to promote the development of domestic independent HPC software.

RA-NLBF： design method of reconfigurable operation unit for stream cipher non-linear Boolean function

Zongren ZHANG, Zibin DAI, Yanjiang LIU, Xiaolei ZHANG

2023, 43(11):  3527-3533.  DOI: 10.11772/j.issn.1001-9081.2022111690

Asbtract ( )   HTML ( )   PDF (1594KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Both the S-box （multiple outputs） in block ciphers and the feedback function in stream ciphers require special Boolean functions to ensure the security of the cipher algorithm. To solve the problems of excessive resource consumption of reconfigurable hardware operation units and low clock frequency caused by Non-Linear Boolean Function （NLBF） in the existing algorithms of stream cipher， a high-efficiency AIC（And-Inverter Cone）-based design scheme for NLBF reconfigurable operation units was proposed， namely RA-NLBF. Based on the theories of cryptography， after analyzing the NLBF characteristics of various stream cipher algorithms and extracting the function features of NLBF including the times of AND terms， the number of AND terms， and the number of input ports， an NLBF simplification method based on the dual-logic hybrid form of “Mixed Polarity Reed-Muller （MPRM）” and “Traditional Boolean function （TB）” was proposed， which reduced the number of NLBF AND terms by 29% and formed an NLBF expression suitable for the AIC. Based on the simplified expression characteristics， such as the distribution of the number of AND terms and the times of AND terms， reconfigurable AIC units and interconnection networks were designed to form the reconfigurable units that can satisfy the NLBF operation in the existing public stream cipher algorithms. The proposed RA-NLBF was verified by logic synthesis based on CMOS 180 nm technology， and the results show that the area of RA-NLBF is 12 949.67 μm2， and the clock frequency reaches 505 MHz， which is a 59.7% reduction in area and a 37.3% increase in clock frequency compared with Reconfigurable Logic Unit for Sequence Cryptographic （RSCLU）， an existing method with the same function.

Genotype imputation algorithm fusing convolution and self-attention mechanism

Jionghuan CHEN, Shengli BAO, Xiaofei WANG, Ruofan LI

2023, 43(11):  3534-3539.  DOI: 10.11772/j.issn.1001-9081.2022111756

Asbtract ( )   HTML ( )   PDF (1678KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Genotype imputation can compensate for the missing due to technical limitations by estimating the sample regions that are not covered in gene sequencing data with imputation， but the existing deep learning-based imputation methods cannot effectively capture the linkage among complete sequence loci， resulting in low overall imputation accuracy and high dispersion of batch sequence imputation accuracy. Therefore， FCSA （Fusing Convolution and Self-Attention）， an imputation method that fuses convolution and self-attention mechanism， was proposed to address the above problems， and two fusion modules were used to form encoder and decoder to construct network model. In the encoder fusion module， a self-attention layer was used to obtain the correlation among complete sequence loci， and the local features were extracted through the convolutional layer after fusing the correlation to global loci. In the decoder fusion module， the local features of the encoded low-dimensional vector were reconstructed by convolution， and the complete sequence was modeled and fused by self-attention layer. The genetic data of multiple species of animals were used for model training， and the comparison and validation were carried out on Dog， Pig and Chicken datasets. The results show that compared to SCDA （Sparse Convolutional Denoising Autoencoders）， AGIC （Autoencoder Genome Imputation and Compression） and U-net， FCSA achieves the highest average imputation accuracy at 10%， 20% and 30% missing rate. Ablation experimental results also show that the design of the two fusion modules is effective in improving the accuracy of genotype imputation.

Network and communications

Joint optimization method for SWIPT edge network based on deep reinforcement learning

Zhe WANG, Qiming WANG, Taoshen LI, Lina GE

2023, 43(11):  3540-3550.  DOI: 10.11772/j.issn.1001-9081.2022111732

Asbtract ( )   HTML ( )   PDF (3553KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Edge Computing （EC） and Simultaneous Wireless Information and Power Transfer （SWIPT） technologies can improve the performance of traditional networks， but they also increase the difficulty and complexity of system decision-making. The system decisions designed by optimization methods often have high computational complexity and are difficult to meet the real-time requirements of the system. Therefore， aiming at Wireless Sensor Network （WSN） assisted by EC and SWIPT， a mathematical model of system energy efficiency optimization was proposed by jointly considering beamforming， computing offloading and power control problems in the network. Then， concerning the non-convex and parameter coupling characteristics of this model， a joint optimization method based on deep reinforcement learning was proposed by designing information interchange process of the system. This method did not need to build an environmental model and adopted a reward function instead of the Critic network for action evaluation， which could reduce the difficulty of decision-making and improve the system real-time performance. Finally， based on the joint optimization method， an Improved Deep Deterministic Policy Gradient （IDDPG） algorithm was designed. Simulation comparisons were made with a variety of optimization algorithms and machine learning algorithms to verify the advantages of the joint optimization method in reducing the computational complexity and improving real-time performance of decision-making.

Incentive mechanism design for hierarchical federated learning based on multi-leader Stackelberg game

Fangxing GENG, Zhuo LI, Xin CHEN

2023, 43(11):  3551-3558.  DOI: 10.11772/j.issn.1001-9081.2022111727

Asbtract ( )   HTML ( )   PDF (2438KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The existence of privacy security and resource consumption issues in hierarchical federated learning reduces the enthusiasm of participants. To encourage a sufficient number of participants to actively participate in learning tasks and address the decision-making problem between multiple mobile devices and multiple edge servers， an incentive mechanism based on multi-leader Stackelberg game was proposed. Firstly， by quantifying the cost-utility of mobile devices and the payment of edge servers， a utility function was constructed， and an optimization problem was defined. Then， the interaction among mobile devices was modeled as an evolutionary game， and the interaction among edge servers was modeled as a non-cooperative game. To solve the optimal edge server selection and pricing strategy， a Multi-round Iterative Edge Server selection algorithm （MIES） and a Gradient Iterative Pricing Algorithm （GIPA） were proposed. The former was used to solve the evolutionary game equilibrium solution among mobile devices， and the latter was used to solve the pricing competition problem among edge servers. Experimental results show that compared with Optimal Pricing Prediction Strategy （OPPS）， Historical Optimal Pricing Strategy （HOPS） and Random Pricing Strategy （RPS）， GIPA can increase the average utility of edge servers by 4.06%， 10.08%， and 31.39% respectively.

Performance analysis of bit error rate on RIS assisted index modulation cooperative system

Chenghao YU, Runhe QIU

2023, 43(11):  3559-3567.  DOI: 10.11772/j.issn.1001-9081.2022101559

Asbtract ( )   HTML ( )   PDF (2563KB) ( )  

Figures and Tables | References | Related Articles | Metrics

For relayed collaborative communications have weak signal of direct paths between the transmitter and the receiver and low Signal-to-Noise Ratio （SNR）， a Reconfigurable Intelligent Surface （RIS） assisted cooperative Index Modulation （IM） system of Decode-and-Forward （DF） relay （RIS-DF-IM） was proposed. In RIS-DF-IM， as smart Access Points （APs）， RISs were adopted as part of the transmitter at the source and relay nodes to perform phase compensation for the reflected channel to maximize the receiving antenna SNR according to the transmission information， and perform IM on multiple antennas of receivers of the relay and destination nodes to improve the spectral efficiency of the system. At the same time， the theoretical union bounds about the Bit Error Rate （BER） of the proposed dual-hop system were solved by using the Moment Generating Function （MGF） method. Besides， a Simplified Pre-greedy Maximum Likelihood （SPML） detector was proposed to reduce the computational complexity by decreasing the number of traversal antenna indexes and simplifying the Maximum Likelihood （ML） decoding criterion formula. Monte Carlo simulation results show that， when the number of RIS elements is 128 and the spatial modulation is adopted， the BER of RIS-DF-IM is about 10 lower than that of the cooperative spatial modulation system where RIS is not connected to the transmitter at the far end； and the BER is dramatically decreased by about 20 compared with the traditional precoded spatial modulation system. Although SPML detector has the BER increased by about 1.4 compared to the Maximum Likelihood （ML） detector， the computational complexity is reduced by a half， achieving an effective balance between BER and complexity.

Computer software technology

Software quality prediction based on back propagation neural network optimized by ant colony optimization algorithm

Jiahao ZHU, Wei ZHENG, Fengyu YANG, Xin FAN, Peng XIAO

2023, 43(11):  3568-3573.  DOI: 10.11772/j.issn.1001-9081.2022101600

Asbtract ( )   HTML ( )   PDF (1715KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Concerning the problems of slow convergence and low accuracy of software quality prediction model based on Back Propagation Neural Network （BPNN）， a Software Quality Prediction method based on BPNN optimized by Ant Colony Optimization algorithm （SQP-ACO-BPNN） was proposed. Firstly， the software quality evaluation factors were selected and a software quality evaluation system was determined. Secondly， BPNN was adopted to build initial software quality prediction model and ACO algorithm was used to determine network structures， initial connection weights and thresholds of network. Then， an evaluation function was given to select the best structure， initial connection weights and thresholds of the network. Finally， the network was trained by BP algorithm， and the final software quality prediction model was obtained. Experimental results of predicting the quality of airborne embedded software show that the accuracy， precision， recall and F1 value of the optimized BPNN model are all improved with faster convergence， which indicates the validity of SQP-ACO-BPNN.

Multimedia computing and computer simulation

Visible and infrared image fusion by preserving gradients and contours

Linkai HAN, Jiangwei YAO, Kunfeng WANG

2023, 43(11):  3574-3578.  DOI: 10.11772/j.issn.1001-9081.2022101553

Asbtract ( )   HTML ( )   PDF (2124KB) ( )  

Figures and Tables | References | Related Articles | Metrics

In order to solve the problems of unclear contours of heat source objects and missing image content in severely exposed regions when visible and infrared images are fused by using basic Laplacian blending， an image fusion algorithm that preserves infrared contours and gradient information was proposed. Firstly， the input image was transformed into color space and denoised by adaptive morphology， and the gradient contrast of the two images and the contour of the highlighted object in the infrared image were taken as the weights of pixel activity information. Secondly， the weights and the input images were decomposed simultaneously， and the weight assignment was adjusted by similarity-based comparison. Finally， the image was reconstructed and the color space was transformed. In subjective evaluation， the proposed algorithm does not produce artifacts and strange colors， and the contours of the heat object in the obtained image is clear. In objective evaluation， the proposed algorithm has an ENtropy （EN） of 7.49， an Edge Intensity （EI） of 74.61， and an Average Gradient （AG） of 7.23， compared with the traditional multi-scale transformation methods （including Non-Subsampled Contourlet Transformation （NSCT） method， the method based on Non-Subsampled Shearlet Transform （NSST） multi-scale entropy） and the latest deep learning method （such as the method combining Residual Network （ResNet） and Zero-phase Component Analysis （ZCA））， it improves EN by 0.10， 0.58 and 0.75， EI by 6.65， 20.35 and 37.35， and AG by 0.73， 2.19 and 3.55； it also achieves a processing speed of 5 frame/s on Intel i5 series computers with low computational complexity.

Small object detection algorithm based on split mixed attention

Qiangqiang QIN, Junguo LIAO, Yixun ZHOU

2023, 43(11):  3579-3586.  DOI: 10.11772/j.issn.1001-9081.2022111660

Asbtract ( )   HTML ( )   PDF (2960KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Focusing on the characteristics of small objects in images， such as less feature information， low percentage， and easy to be influenced by the environment， a small object detection algorithm based on split mixed attention was proposed， namely SMAM-YOLO. Firstly， by combining Channel Attention （CA） and Spatial Attention （SA）， as well as recombining the connection structures， a Mixed Attention Module （MAM） was proposed to enhance the model’s representation of small object features in spatial dimension. Secondly， according to the different influence of receptive fields with different sizes on the object， a Split Mixed Attention Module （SMAM） was proposed to adaptively adjust the size of the receptive field according to the scale of the input feature map， and the mixed attention was used to enhance the ability to capture small object feature information in different branches. Finally， the core residual module in YOLOv5 was improved by using SMAM， and a feature extraction module CSMAM was proposed on the basis of CSPNet （Cross Stage Partial Network） and SMAM， and the additional computational overhead of CSMAM can be ignored. Experimental results on TinyPerson dataset show that compared with the baseline algorithm YOLOv5s， when the Intersection over Union （IoU） threshold is 0.5， the mean Average Precision （mAP₅₀） of SMAM-YOLO algorithm is improved by 4.15 percentage points， and the detection speed reaches 74 frame/s. In addition， compared with some existing mainstream small object detection models， SMAM-YOLO algorithm improves the mAP₅₀ by 1.46 - 6.84 percentage points on average， and it can meet the requirements of real-time detection.

Multi-scale ship detection based on adaptive feature fusion in complex scenes

Fang LUO, Yang LIU, G. T. S HO

2023, 43(11):  3587-3593.  DOI: 10.11772/j.issn.1001-9081.2022101593

Asbtract ( )   HTML ( )   PDF (1778KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Under the influence of complex weather such as typhoon， heavy fog， rain and snow， as well as occlusions and scale changes， the existing ship detection methods have the problems of false detection and missed detection. In order to solve the above complex scene problems， based on YOLOX-S model， a multi-scale ship detection method based on adaptive feature fusion was proposed. Firstly， a feature augmentation module was introduced into the backbone feature extraction network to suppress the interference of complex background noise on ship feature extraction. Then， considering the problem of deep and shallow feature fusion proportion， an adaptive feature fusion module was designed to make full use of deep and shallow features， thereby improving the multi-scale ship detection ability of the model. Finally， in the detection head network， the detection head was decoupled and an adaptive multi-task loss function was introduced to balance classification tasks and regression tasks， thereby improving the multi-scale ship detection robustness of the model. Experimental results show that the detection mean Average Precision （mAP） of the proposed method on the public ship detection datasets SeaShips and McShips is 97.43% and 96.10%， respectively. The detection speed of the proposed method reaches 189 frames per second， which meets the requirements of real-time detection， demonstrating that the proposed method achieves high-precision detection of multi-scale ship targets even in complex scenes.

Trajectory prediction of sea targets based on geodetic distance similarity calculation

Yijian ZHAO, Li LIN, Qianqian WANG, Peng WEN, Dong YANG

2023, 43(11):  3594-3598.  DOI: 10.11772/j.issn.1001-9081.2022101639

Asbtract ( )   HTML ( )   PDF (1803KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The existing similarity-based moving target trajectory prediction algorithms are generally classified according to the spatial-temporal characteristics of the data， and the characteristics of the algorithms themselves cannot be reflected. Therefore， a classification method based on algorithm characteristics was proposed. The calculation of the distances between two points is required for the trajectory similarity algorithms to carry out the subsequent calculations， however， the commonly used Euclidean Distance （ED） is only applicable to the problem of moving targets in a small region. A method of similarity calculation using geodetic distance instead of ED was proposed for the trajectory prediction of sea targets moving in a large region. Firstly， the trajectory data were preprocessed and segmented. Then， the discrete Fréchet Distance （FD） was adopted as similarity measure. Finally， synthetic and real data were used to test. Experimental results indicate that when sea targets move in a large region， the ED-based algorithm may gain incorrect prediction results， while the geodetic distance-based algorithm can output correct trajectory prediction.

Segmentation of breast nodules in ultrasound images based on multi-scale and cross-spatial fusion

Xin ZHAO, Qianqian ZHU, Cong ZHAO, Jialing WU

2023, 43(11):  3599-3606.  DOI: 10.11772/j.issn.1001-9081.2022111673

Asbtract ( )   HTML ( )   PDF (3808KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Accurate breast nodule segmentation in ultrasound images is very challenging due to the low resolution and noise of ultrasound imaging， as well as the complexity and variability of the shape and texture of nodules， therefore， an end-to-end automatic segmentation method of breast nodules in ultrasound images based on multi-scale and cross-spatial feature fusion was proposed. Firstly， a Multi-scale Feature Extraction and Fusion （MFEF） module was designed to enable the network to have multi-scale feature extraction ability by fusing four convolutional paths with different receptive fields. Then， for the multi-scale observation and information filtering of high-level semantic information， a Scale-aware Feature Aggregation （SFA） module was used at the bottleneck layer to enhance the deep feature extraction ability in the encoding stage. Besides， a Cross-spatial Residual Fusion （CRF） module was designed and applied to the skip connection between the encoder and decoder to fuse information among different encoding layers in a cross-spatial way and implement information complementarity between different encoding layers， further extract information features of encoding layer and narrow the difference between peer layers of encoder and decoder to better compensate for the information loss in the decoding stage. Experimental results on a public ultrasound breast nodule dataset show that the proposed method achieves DICE coefficient of 0.888， which is 0.033 to 0.094 higher than those of the mainstream deep learning segmentation models UNet， AttUNet， ResUNet++ and SKUNet， and is 0.001 to 0.068 higher than those of the improved models such as CF²-Net， Estan， FS-UNet and SMU-Net in the same dataset. The subjective visualization of the segmentation result of the proposed method is closest to the gold standards provided by experts， verifying that the proposed mehtod can segment the breast nodule area more accurately.

Speech enhancement algorithm based on multi-scale ladder-type time-frequency Conformer GAN

Yutang JIN, Yisong WANG, Lihui WANG, Pengli ZHAO

2023, 43(11):  3607-3615.  DOI: 10.11772/j.issn.1001-9081.2022111734

Asbtract ( )   HTML ( )   PDF (4515KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Aiming at the problem of artificial artifacts due to phase disorder in frequency-domain speech enhancement algorithms， which limits the denoising performance and decreases the speech quality， a speech enhancement algorithm based on Multi-Scale Ladder-type Time-Frequency Conformer Generative Adversarial Network （MSLTF-CMGAN） was proposed. Taking the real part， imaginary part and magnitude spectrum of the speech spectrogram as input， the generator first learned the local and global feature dependencies between temporal and frequency domains by using time-frequency Conformer at multiple scales. Secondly， the Mask Decoder branch was used to learn the amplitude mask， and the Complex Decoder branch was directly used to learn the clean spectrogram， and the outputs of the two decoder branches were fused to obtain the reconstructed speech. Finally， the metric discriminator was used to judge the scores of speech evaluation metrics， and high-quality speech was generated by the generator through minimax training. Comparison experiments with various types of speech enhancement models were conducted on the public dataset VoiceBank+Demand by subjective evaluation Mean Opinion Score （MOS） and objective evaluation metrics.Experimental results show that compared with current state-of-the-art speech enhancement method CMGAN （Comformer-based MetricGAN）， MSLTF-CMGAN improves MOS prediction of the signal distortion （CSIG） and MOS predictor of intrusiveness of background noise （CBAK） by 0.04 and 0.07 respectively， even though its Perceptual Evaluation of Speech Quality （PESQ） and MOS prediction of the overall effect （COVL） are slightly lower than that of CMGAN， it still outperforms other comparison models in several subjective and objective speech evaluation metrics.

Frontier and comprehensive applications

Multi-site wind speed prediction based on graph dynamic attention network

Bolu LI, Li WU, Xiaoying WANG, Jianqiang HUANG, Tengfei CAO

2023, 43(11):  3616-3624.  DOI: 10.11772/j.issn.1001-9081.2022111749

Asbtract ( )   HTML ( )   PDF (4716KB) ( )  

Figures and Tables | References | Related Articles | Metrics

The task of spatio-temporal sequence prediction has a wide range of applications in the fields such as transportation， meteorology and smart city. It is necessary to learn the spatio-temporal characteristics of different data with the combination of external factors such as precipitation and temperature when making station wind speed predictions， which is one of the main tasks in meteorological forecasting. The irregular distribution of meteorological stations and the inherent intermittency of the wind itself bring the challenge of achieving wind speed prediction with high accuracy. In order to consider the influence of multi-site spatial distribution on wind speed to obtain accurate and reliable prediction results， a Graph-based Dynamic Switch-Attention Network （Graph-DSAN） wind speed prediction model was proposed. Firstly， the distances between different sites were used to reconstruct the connection of them. Secondly， the process of local sampling was used to model adjacency matrices of different sampling sizes to achieve the aggregation and transmission of the information between neighbor nodes during the graph convolution process. Thirdly， the results of the graph convolution processed by Spatio-Temporal Position Encoding （STPE） were fed into the Dynamic Attention Encoder （DAE） and Switch-Attention Decoder （SAD） for dynamic attention computation to extract the spatio-temporal correlations. Finally， a multi-step prediction was formed by using autoregression. In experiments on wind speed prediction on 15 sites data in New York State， the designed model was compared with ConvLSTM， Graph Multi-Attention Network （GMAN）， Spatio-Temporal Graph Convolutional Network （STGCN）， Dynamic Switch-Attention Network （DSAN） and Spatial-Temporal Dynamic Network （STDN）. The results show that the Root Mean Square Error （RMSE） of 12 h prediction of Graph-DSAN model is reduced by 28.2%， 6.9%， 27.7%， 14.4% and 8.9% respectively， verifying the accuracy of Graph-DSAN in wind speed prediction.

Accident prediction model fusing heterogeneous traffic situations

Bo YANG, Zongtao DUAN, Pengfei ZUO, Yuanyuan XIAO, Yilin WANG

2023, 43(11):  3625-3631.  DOI: 10.11772/j.issn.1001-9081.2022101619

Asbtract ( )   HTML ( )   PDF (2056KB) ( )  

Figures and Tables | References | Related Articles | Metrics

To address the problems of limited information expression， imbalance， and dynamic spatio-temporal characteristics of accident data， an accident prediction model fusing heterogeneous traffic situations was proposed. In which， the semantic enhancement was completed by the spatio-temporal state aggregation module through traffic events and weather features representing dynamic traffic situations， and the historical multi-period spatio-temporal states of four types of regions （single region， adjacent region， similar region， and global region） were aggregated； the dynamic local and global spatio-temporal characteristics of accident data were captured by the spatio-temporal relation capture module from both micro- and macro-perspectives； and the multi-region and multi-angle spatio-temporal states were further fused by the spatio-temporal data fusion module， and the accident prediction task in the next period was realized. Experimental results on five city datasets of US-Accident demonstrate that the average F1-scores of the proposed model for accident， non-accident， and weighted average samples are 85.6%， 86.4%， and 86.6% respectively， which are improved by 14.4%， 5.6%， and 9.3% in the three metrics compared to the traditional Feedforward Neural Network （FNN）， indicating that the proposed model can effectively suppresses the influence of accident data imbalance on experimental results. Constructing an efficient accident prediction model helps to analyze the safety situation of road traffic， reduce the occurrence of traffic accidents and improve the traffic safety.

Situation prediction of flight conflict network based on online fuzzy least squares support vector machine with optimal training set

Xiangxi WEN, Yating PENG, Kexin BI, Yuming HENG, Minggong WU

2023, 43(11):  3632-3640.  DOI: 10.11772/j.issn.1001-9081.2022101605

Asbtract ( )   HTML ( )   PDF (3403KB) ( )  

Figures and Tables | References | Related Articles | Metrics

Concerning the periodicity and time-varying characteristics of air traffic system operation， a flight conflict network situation prediction method based on Optimal Training Set Online Fuzzy-Least Squares Support Vector Machine （OTSOF-LSSVM） was proposed by combining complex network theory and fuzzy Least Squares Support Vector Machine （LSSVM）. Firstly， a flight conflict network model was constructed based on the three-dimensional velocity obstacle method， and conflicts were judged according to the positions， headings and velocities of the aircrafts. Then， the evolution time series of topology indicators of flight conflict network were analyzed to obtain the optimal training set which consisted of samples related to the predicted moment in time and distance. Finally， a prediction model was obtained by online fuzzy LSSVM training， and the idea of block matrix was used to simplify the updating process and improve the efficiency of the algorithm. Experimental results show that the proposed method can quickly and accurately predict the air situation， provide reference for controllers to master the development of air traffic， and assist the pre-deployment of conflicts.

UAV cluster cooperative combat decision-making method based on deep reinforcement learning

Lin ZHAO, Ke LYU, Jing GUO, Chen HONG, Xiancai XIANG, Jian XUE, Yong WANG

2023, 43(11):  3641-3646.  DOI: 10.11772/j.issn.1001-9081.2022101511

Asbtract ( )   HTML ( )   PDF (2944KB) ( )  

Figures and Tables | References | Related Articles | Metrics

When the Unmanned Aerial Vehicle （UAV） cluster attacks ground targets， it will be divided into two formations： a strike UAV cluster that attacks the targets and a auxiliary UAV cluster that pins down the enemy. When auxiliary UAVs choose the action strategy of aggressive attack or saving strength， the mission scenario is similar to a public goods game where the benefits to the cooperator are less than those to the betrayer. Based on this， a decision method for cooperative combat of UAV clusters based on deep reinforcement learning was proposed. First， by building a public goods game based UAV cluster combat model， the interest conflict problem between individual and group in cooperation of intelligent UAV clusters was simulated. Then， Muti-Agent Deep Deterministic Policy Gradient （MADDPG） algorithm was used to solve the most reasonable combat decision of the auxiliary UAV cluster to achieve cluster victory with minimum loss cost. Training and experiments were performed under conditions of different numbers of UAV. The results show that compared to the training effects of two algorithms — IDQN （Independent Deep Q-Network） and ID3QN （Imitative Dueling Double Deep Q-Network）， the proposed algorithm has the best convergence， its winning rate can reach 100% with four auxiliary UAVs， and it also significantly outperforms the comparison algorithms with other UAV numbers.