Content of Network, Communication and Security in our journal

        Published in last 1 year |  In last 2 years |  In last 3 years |  All
    Please wait a minute...
    For Selected: Toggle Thumbnails
    Searchable Encrypted Data Sharing Scheme Based on Inverted Index
    LIU Wei, BAI Xiaodan, SHE Wei, SONG Xuan, TIAN Zhao
    Computer Engineering and Applications    2023, 59 (10): 270-279.   DOI: 10.3778/j.issn.1002-8331.2201-0159
    Abstract26)      PDF(pc) (819KB)(11)       Save
    For the problems such as unreliable data sharing, data attacked and low efficient of retrieval ciphertext, this paper proposes a searchable encrypted data sharing scheme based on inverted index. Firstly, the double chain structure based on private blockchain and consortium blockchain is adopted to store data and share data. Secondly, a new inverted index structure is designed to prevent the sensitive data from being attacked. Finally, a ciphertext search algorithm based on the new index structure is proposed, which uses the searchable encryption technology and submits the trapdoor to inverted index structure, to realize ciphertext retrieval. The experimental results show that the proposed scheme can guarantee data security and improve retrieval efficiency.
    Reference | Related Articles | Metrics
    Three-Factor Anonymous Authentication Protocol in Telecare Medicine Information System
    LI Yi, TIAN Yuling
    Computer Engineering and Applications    2023, 59 (10): 280-287.   DOI: 10.3778/j.issn.1002-8331.2201-0238
    Abstract31)      PDF(pc) (597KB)(15)       Save
    Telecare medicine information system provides seamless medical information transfer and timely sharing for specific patients, and user authentication protocol is one of important technologies to ensure security and privacy in telecare medicine information system. This paper analyzes the security of Dharminder’s RSA-based authentication schemes for authorized access to healthcare services, and points out that the scheme cannot resist key compromise impersonation attack, denial of service attack, and cannot provide forward security. To this end, this paper improves the scheme of Dharminder et al., and proposes a three-factor anonymous authentication protocol based on elliptic curve cryptography. The proposed protocol retains the general flow of Dharminder et al., replaces the RSA method with elliptic curve algorithm and symmetric encryption/decryption algorithm, and corrects the errors in the last step of the authentication process of Dharminder et al. A formal security analysis is performed for the proposed protocol using BAN(Burrows-Abadi-Needham) logic, and a formal security verification is implemented using AVISPA(automated validation of Internet security-sensitive protocols and applications). The results show that the proposed protocol can resist various malicious attacks. Besides, performance analysis using MIRACL(multiprecision integer and rational arithmetic C/C++library) shows that the new protocol has performance advantages over other recent schemes.
    Reference | Related Articles | Metrics
    DP-IMKP:Data Publishing Protection Method for Personalized Differential Privacy
    ZHANG Xing, ZHANG Xing, WANG Qingyang
    Computer Engineering and Applications    2023, 59 (10): 288-298.   DOI: 10.3778/j.issn.1002-8331.2201-0457
    Abstract25)      PDF(pc) (718KB)(24)       Save
    Differential privacy is widely used to solve the problem of privacy protection in data publishing because of its powerful privacy guarantee. However, the data protected by differential privacy are injected with a lot of noise, which reduces the data utility. In addition, in the existing methods, there are few research results on privacy protection published for mixed attribute datasets and unreasonable allocation of privacy budget. Therefore, this paper proposes a differential privacy mixed attribute data publishing method based on personalized privacy budget allocation(DP-IMKP). Firstly, based on the correlation between mutual information and attributes, a classification strategy for sensitive attributes is proposed to quantify the importance of each attribute, and match the corresponding privacy protection degree for different levels of attributes. Secondly, combined with the optimal matching theory, a bipartite graph between privacy budget and sensitive attributes is constructed, the reasonable privacy budget is allocated for sensitive attributes at all levels. Combined with the idea of information entropy and density optimization, the selection of initial center and the measurement method of dissimilarity in classical [k]-prototype algorithm are improved, and the privacy budget allocated by each sensitive attribute is used to implement differential privacy protection for the clustering center value to prevent the disclosure of private data information. Experimental results show that compared with similar methods, DP-IMKP has obvious advantages in improving data utility and reducing data leakage risk.
    Reference | Related Articles | Metrics
    Control Flow Obfuscation Scheme for LLVM Intermediate Languages
    LI Chengyang, HUANG Tianbo, CHEN Xiarun, WEN Weiping
    Computer Engineering and Applications    2023, 59 (8): 263-269.   DOI: 10.3778/j.issn.1002-8331.2112-0035
    Abstract57)      PDF(pc) (528KB)(19)       Save
    Software security issues are becoming more prominent in the post-epidemic era, and code obfuscation as a mature protection scheme provides the possibility of cross-platform use with the help of LLVM. However, LLVM-based control flow obfuscation algorithms are limited in terms of protection strength, on the one hand, the existing algorithm model is immutable and lacks structural innovation. On the other hand, the obfuscation processing does not take into account the fact that attackers can base on the basic block. Therefore, two algorithms are proposed:firstly, nested switch obfuscation, which breaks the inherent flat processing model and enhances the hiding of the hopping amount by reconstructing the switch structure internally; secondly, indegree obfuscation, which adds an anti-entry degree analysis strategy to the false control flow to circumvent the false block by changing the indegree of the false block. The results show that the obfuscation method can further reduce 58.67% of the basic block similarity and increase 64.44% of the jump instructions compared to the existing control-flow obfuscation scheme within 1.5 times the temporal overhead.
    Reference | Related Articles | Metrics
    Blockchain-Based Attribute Proxy Re-Encryption Data Sharing Scheme
    ZHAI Sheping, TONG Tong, BAI Xifang
    Computer Engineering and Applications    2023, 59 (8): 270-279.   DOI: 10.3778/j.issn.1002-8331.2205-0115
    Abstract74)      PDF(pc) (674KB)(29)       Save
    The rapid growth of data has put forward new requirements for data management methods. Aiming at the problems of centralized data storage and the difficulty of data sharing in traditional data management methods, this paper proposes a data security sharing scheme based on blockchain and attribute agent re-encryption. Firstly, to solve the problem of secure data sharing, the original data after symmetric encryption is stored in ciphertext chain, the key and index information is stored in index chain, and the secure sharing of key information on the index chain is completed by re-encrypting the data attributes proxy. It meets the user’s demand for fine-grained data access control, realizes multi-user decryption authority authorization, and ensures the security of data sharing process. Then, the distributed key generation method is designed and updated periodically to avoid the centralized hosting problem and disclosure risk of encryption key management. Finally, through the comparison and simulation with the existing attribute agent re-encrypted data sharing scheme, the security and efficiency of this scheme are verified.
    Reference | Related Articles | Metrics
    Defense Mechanism to Solve Eclipse Attack of POW Consensus
    WEI Zizuan, WANG Xin, YU Dan, MA Yao, CHEN Yongle
    Computer Engineering and Applications    2023, 59 (8): 280-287.   DOI: 10.3778/j.issn.1002-8331.2112-0552
    Abstract58)      PDF(pc) (680KB)(21)       Save
    The eclipse attack is a malicious attack method in the blockchain system. Attackers achieve the purpose of deception and control by monopolizing the network of victim nodes. It is known from actual  blockchain applications that there is no good way to defend against eclipse attacks. Therefore, in order to solve this problem, a dynamic defense model for solving eclipse attacks based on the POW consensus blockchain system is designed, which is based on mutual evaluation mechanism between nodes. Specifically, based on kademlia algorithm, it evaluates and stores a level for every node on the result of mutual evaluation between clients to design a level evaluation mechanism. The nodes select normal peer nodes as their neighbors and avoid malicious nodes according to the relation between level value and level limitation value. Experiments have proved that this method can effectively resist eclipse attacks. This method improves a series of defense strategies previously proposed. It does not need to change the protocol and network of the blockchain system. It can successfully resist eclipse attacks with a high probability, and the overhead  generated by the model is also very small, which is very suitable for actual deployment.
    Reference | Related Articles | Metrics
    Optimization of Charging Scheduling Under Multi-Node Partial Charging Model
    LIU Jingxiang, XU Wenzheng
    Computer Engineering and Applications    2023, 59 (6): 251-257.   DOI: 10.3778/j.issn.1002-8331.2111-0410
    Abstract51)      PDF(pc) (624KB)(26)       Save
    In the wireless rechargeable sensor network, the mobile charger adopts the multi-node partial charging model to reciprocate during the charging scheduling process, which leads to the increase of the charging duration. For this reason, this paper proposes a novel multi-node partial charging model, which globally optimizes the charging duration of the charger at each charging position, ensuring that each energy-critical sensor is fully charged. Meanwhile, the AlgMinTime algorithm is designed for path planning to determine the charging tour of the charger and the charging duration corresponding to the charging position, such that the total charging scheduling duration in the tour is minimized. Finally, the performance of the proposed algorithm is evaluated through simulation experiments. Experimental results show that the average tour duration of charging scheduling by the proposed algorithm is 9.8% shorter than that of the SOTA algorithm.
    Reference | Related Articles | Metrics
    Smart Contract Election Scheme with Short Linkable Ring Signatures
    WANG Jiechang, ZHANG Ping, DUAN Ying, LIU Yuling, WANG Xiaowei
    Computer Engineering and Applications    2023, 59 (6): 258-267.   DOI: 10.3778/j.issn.1002-8331.2111-0405
    Abstract72)      PDF(pc) (637KB)(27)       Save
    At present, using linkable ring signatures or one-time ring signatures, some electronic election protocols protect the privacy of voters, and prevent voting repeatedly. However, the size of the signature increases with the number of voters, while the size of the short linkable ring signature remains constant, but existing short linkable ring signatures are inefficient. To solve these problems, using accumulator and signatures based on proofs of knowledge, new efficient short linkable ring signatures are constructed, and combining signatures with the anonymous address and IPFS, a new smart contract election scheme is proposed. The operations of election setup, voting and counting phases are designed respectively. The unforgeability, linkability, anonymity, privacy, public verifiability and operational efficiency of the scheme are proved and analyzed. Finally, an experimental simulation evaluation is carried out, and the results show that with the increase of voters, the signature size and Gas cost of the ballot are less and constant, and the generation time and verification time are less and increase slowly.
    Reference | Related Articles | Metrics
    Side Channel Attack Fused with CNN_LSTM
    PENG Pei, ZHANG Meiling, ZHENG Dong
    Computer Engineering and Applications    2023, 59 (6): 268-276.   DOI: 10.3778/j.issn.1002-8331.2111-0511
    Abstract63)      PDF(pc) (769KB)(34)       Save
    Based on the application of deep learning in side-channel attacks, the AES algorithm is first implemented in the Chipwhisperer platform, the corresponding energy trace is measured during its encryption process, and then the CPA technology is used to analyze the location of the point of interest, and model training is made for the point of interest. On the three network models convolutional neural network(CNN), long-short-term memory network(LSTM) and CNN_LSTM hybrid model, combined with data preprocessing technology to train synchronous and asynchronous energy traces. The experimental results show that the accuracy of the three models in the synchronous state is equivalent. In addition, when the asynchronous data is gradually increased while ensuring the model training parameters remain unchanged, the accuracy of the training set and test set of the three models is decreasing, but the declining speed of the new hybrid model is the slowest. When the experimental asynchronous number is increased to 50, the accuracy can still be guaranteed to be above 90%, that is, the correct key can be recovered with almost one energy trace. Therefore, the CNN_LSTM model can better adapt to the situation where the energy trace occurs asynchronously.
    Reference | Related Articles | Metrics
    Design and Implementation of Intelligent Interference Avoidance System Based on NS3-gym Framework
    CHEN Haitao, GONG Guangwei, ZHANG Jiao, ZHAO Haitao, XIONG Jun, WEI Jibo, ZHAN Dechuan
    Computer Engineering and Applications    2023, 59 (4): 252-260.   DOI: 10.3778/j.issn.1002-8331.2111-0074
    Abstract114)      PDF(pc) (850KB)(48)       Save
    The agents need to interact with the environment to learn the interference features and optimize the anti-interference strategies, when the multi-agent wireless communication networks are subject to the external malicious interference. It is necessary to design an intelligent interference avoidance simulation platform in order to effectively simulate and verify the learning interaction between agent and external interference environment. An intelligent interference avoidance system based on NS3-gym framework is proposed. NS3 implements a network simulation scenario and shares the sensed network state data as the input of agent. Agent provides interference avoidance strategy via learning and analyzing the input data, then returns it to NS3 for anti-interference strategy deployment through the interaction between gym and NS3. NS3-gym framework provides an interface for information exchange between NS3 and OpenAI gym. The simulation platform is built under Ubuntu20.04 system. The performance of Q-learning algorithm and WoLF-PHC algorithm is verified in four interference scenarios:sweep interference, greedy random strategy interference, follow interference and random interference, respectively. The simulation results indicate that the proposed system architecture and the simulation platform are correct and efficient.
    Reference | Related Articles | Metrics
    Trust Management Mechanism of SDVN for Spatio-Temporal Feature Evaluation
    CAI Peijing, LUO Wei, HU Yulin, WU Jing
    Computer Engineering and Applications    2023, 59 (4): 261-268.   DOI: 10.3778/j.issn.1002-8331.2111-0014
    Abstract62)      PDF(pc) (620KB)(20)       Save
    Vehicular ad-hoc network(VANET) has broad application prospects in improving traffic efficiency and solving traffic safety, but its development also brings challenges. With the increase of intelligent vehicles, the existing architecture cannot meet the rapid growth of vehicle demand, and the openness of VANET also encourages the vicious competition of vehicles. This paper uses software-defined vehicle network(SDVN) architecture to improve the network performance of the Internet of vehicles, and designs a trust evaluation and trust management mechanism based on this architecture. The trust evaluation method considers the temporal and spatial characteristics of vehicles to limit and standardize vehicle behaviors and improve the accuracy of vehicle malicious behavior detection. In addition, this mechanism designs a trust management method based on Hyperledger Fabric, which realizes the data security of vehicles with the help of blockchain, and improves the efficiency of updating and querying trust values. Experimental analysis shows that the proposed trust management mechanism can ensure message consistency and security in SDVN network, shorten the consensus time, and improve the overall efficiency of trust management.
    Reference | Related Articles | Metrics
    Adversarial Attacks for Object Detection Based on Region of Interest of Feature Maps
    WANG Yekui, CAO Tieyong, ZHENG Yunfei, FANG Zheng, WANG Yang , LIU Yajiu, FU Bingyang, CHEN Lei
    Computer Engineering and Applications    2023, 59 (2): 261-270.   DOI: 10.3778/j.issn.1002-8331.2206-0184
    Abstract72)      PDF(pc) (4265KB)(56)       Save
    Object detection is widely used in the fields of unmanned driving, monitoring and security. However, it is found that the object detection system is vulnerable to the impact of adversarial examples, resulting in performance degradation, which poses a great danger to its application safety. Most of the adversarial examples for object detection are only designed for a certain type of object detection model, and their transferability is weakly. In order to solve the above problem, based on the generative adversarial networks, an adversarial examples method for object detection is proposed. In this method, a position regression attack loss is designed for the non-maximum suppression mechanism, which is commonly used in the detection model and the key regions predicted by the detection model. Through this loss, the non-maximum suppression mechanism of the model is invalid, and the generated region proposals are guided to deviate from the predicted key regions, resulting in the failure of the model prediction. The experimental results on VOC dataset show that the proposed method can effectively attack object detection models, such as Faster-RCNN, SSD300, SSD512, RetinaNet, YOLOv5, One-Net, etc., which improve the transferability of the adversarial examples for object detection.
    Reference | Related Articles | Metrics
    Android Malicious Application Family Classification Model Incorporating MAML and CBAM
    SU Qing, LIN Jiarui, HUANG Haibin, HUANG Jianfeng
    Computer Engineering and Applications    2023, 59 (2): 271-279.   DOI: 10.3778/j.issn.1002-8331.2110-0492
    Abstract51)      PDF(pc) (3723KB)(28)       Save
    To meet the demand for fast detection of emerging Android malicious application families, it proposes a classification model MAML-CAS that fuses MAML(model-agnostic meta-learning) and CBAM(convolutional block attention module) for Android malicious application families. The DEX files in the sample set of Android malicious apps are visualized as grayscale maps and a task set is constructed; then two convolutional neural networks with equal structure are designed as the base learner and meta-learner respectively by fusing CBAM, which can enhance the key feature representation in both channel and space dimensions while automatically extracting the sample features in the task set; then the meta-learning method is used to MAML is used to train the two learners, where the base learner learns the attributes of a specific malicious family classification task and the meta-learner learns the commonalities of different tasks; after the training of the two learners is completed, MAML-CAS will obtain the initialization parameters, and when faced with a new Android malicious app family classification task, no retraining is required, and only a small number of samples are needed for fast iteration; finally, using the trained base learner is finally used to extract Android malicious app family features and perform malicious family classification using SVM. The experimental results show that the MAML-CAS model has good detection effect on emerging small-sample Android malicious application families, with faster detection speed and better stability.
    Reference | Related Articles | Metrics
    Design of DAPGD of Adversarial Attack Algorithm Against Deepfake
    QIU Haoxuan, DU Yanhui, LU Tianliang
    Computer Engineering and Applications    2022, 58 (24): 97-106.   DOI: 10.3778/j.issn.1002-8331.2110-0026
    Abstract50)      PDF(pc) (2541KB)(39)       Save
    An improved adversarial example generation algorithm, dynamic APGD(DAPGD) is proposed to protect the images from the tampering of deepfake models. Adversarial examples generated by DAPGD make the output of the deepfake models significantly distorted so that the forged images cannot be generated effectively. DAPGD uses the idea of the adaptive decay learning rate, which can accelerate the algorithm convergence and improve the quality of adversarial examples. Meanwhile, the checkpoint for decaying the learning rate is dynamically set to address the problem that APGD tends to miss the best time to decay the learning rate. It can play the role of decaying the learning rate more thoroughly. Finally, as the loss function is unstable due to the use of random parameters in deepfake models, the local early stopping mechanism of APGD is eliminated to improve the effectiveness and speed of the algorithm. DAPGD adversarial attack experiments are conducted for three mainstream deepfake models and compared with the original algorithm and other algorithms. The results show that the adversarial examples generated by DAPGD can achieve better results in both output distortion size and attack success rate, and can interfere with deepfake models forgery images more effectively.
    Reference | Related Articles | Metrics
    Level-Based Learning Swarm Optimization Algorithm for Resource Scheduling in Edge Computing
    HU Xiaomin, CHEN Zhentian, LI Min
    Computer Engineering and Applications    2022, 58 (24): 107-115.   DOI: 10.3778/j.issn.1002-8331.2110-0060
    Abstract46)      PDF(pc) (1505KB)(54)       Save
    Due to problems such as slow CPU operation speed and low battery capacity, smart mobile devices themselves cannot execute applications with large computing requirements. It needs the help of edge computing technology to reduce the requirements of programs on mobile device hardware. However, by transferring the computing burden from mobile devices to edge computing servers, additional transmission energy and server computing power consumption are needed. By considering the four factors influencing the energy consumptions of mobile devices, servers, and data transmission, i.e. the computing speed, data downloading power consumption, the proportion of data offloading and the remain network bandwidth, this paper proposes a level-based learning particle swarm optimization algorithm to optimize the values of these four parameters for each mobile device and more reasonably allocate of computing resources for minimizing the total energy consumption. While modeling of computing resources, the constraints of maximum energy consumption, computing cycle, storage, bandwidth and delay are also considered. Compared with other algorithms, experiments show that the level-based learning particle swarm optimization algorithm can obtain the optimal resource scheduling solution which meet the constrains and with lower energy consumption more quickly.
    Reference | Related Articles | Metrics
    Blockchain Data Synchronization Method Based on NDN
    MA Hongqiao, YANG Wenzhong, KANG Peng
    Computer Engineering and Applications    2022, 58 (23): 117-125.   DOI: 10.3778/j.issn.1002-8331.2204-0060
    Abstract72)      PDF(pc) (1173KB)(35)       Save
    In recent years, blockchain technology has received extensive attention in various fields and achieved phased success. However, its own defects limit its development in important fields. This is because blockchain technology makes its requests for the same data content independent through the communication mode of P2P network based on TCP architecture, resulting in a large amount of redundant traffic during data synchronization, thus occupying a high bandwidth, resulting in a high data transmission delay, and ultimately reducing the overall performance of the blockchain. Therefore, the “NDN+Blockchain” structure is proposed in order to effectively solve the problem of low data transmission efficiency of blockchain and expand the technical advantages of blockchain with the help of the inherent advantages of named data networking(NDN). Through the analysis of simulation experiments, it is concluded that the “NDN+blockchain” structure has a good performance effect in bandwidth occupation and data synchronization delay compared with the traditional TCP blockchain.
    Reference | Related Articles | Metrics
    Research on Algorithm Based on Spectral Continuity in Elastic Optical Networks
    CHEN Bingjun, ZHANG Ning, YANG Yansong, CHEN Xiaodan
    Computer Engineering and Applications    2022, 58 (23): 126-131.   DOI: 10.3778/j.issn.1002-8331.2109-0503
    Abstract51)      PDF(pc) (1333KB)(24)       Save
    Aiming at the problems of limited available wavelength resources and low spectrum utilization in optical networks, a spectrum continuity degree perception algorithm(DP) based on KSP algorithm is proposed. In terms of routing, the KSP algorithm is used to obtain different path lengths between source and destination nodes, and different paths are allocated according to the amount of spectrum resources required for service requests. In terms of spectrum allocation, the algorithm will perceive the spectrum continuity of each link and minimize the spectrum fragments of each link on the service allocated path. The simulation results show that the proposed algorithm can reduce the spectrum blocking probability and improve the spectrum utilization compared with the traditional shortest path RMSA algorithm.
    Reference | Related Articles | Metrics
    Master-Slave Blockchain Fault-Tolerant Heterogeneous Cross-Domain Identity Authentication Scheme
    ZHAO Ping, WANG Ze, LI Fang, SUN Shimin
    Computer Engineering and Applications    2022, 58 (22): 79-88.   DOI: 10.3778/j.issn.1002-8331.2109-0300
    Abstract65)      PDF(pc) (1015KB)(49)       Save
    Heterogeneous cross-domain identity authentication is a technology that performs identity confirmation and security information exchange for nodes in different institutional trust domains. The existing authentication schemes mainly have issues such as single-point attack risk, complex authentication. This paper designs a master-slave blockchain identity authentication model and a hierarchical Byzantine fault-tolerant algorithm for matching. Through the step-by-step and phase-by-phase consensus of the master-slave chain, the number of nodes participating in the consensus is reduced. The unique function nodes of the PKI system and the CL-PKC system correspond to the master-slave chain nodes. On the premise of not changing the function of the original trusted domain node, the hash value of the blockchain certificate is used to efficiently transmit trust, and the authentication is optimized. The process realizes two-way heterogeneous cross-domain identity authentication. In the end, through the simulation experiment and the analysis of security and performance, the result shows that compared the mentioned scheme with others, consensus efficiency and fault tolerance are improved, and communication overhead is reduced while ensuring secure communication.
    Reference | Related Articles | Metrics
    Node Election Scheme for Transaction Overload Processing in State Sharding
    QIN Wenhui, LI Zhihuai, MA Hongcheng
    Computer Engineering and Applications    2022, 58 (22): 89-100.   DOI: 10.3778/j.issn.1002-8331.2109-0341
    Abstract63)      PDF(pc) (1155KB)(40)       Save
    Sharding technology is the most effective blockchain capacity expansion scheme to achieve high performance without reducing the degree of decentralization. Only state sharding can fundamentally solve the blockchain resource bottleneck. Aiming at the problem of transaction overload caused by sharding technology random allocation transactions, a multi-round node election equalization verification scheme for transactions overload processing under state sharding constraints is proposed. The transactions verification in the shardings is divided into multiple rounds. After each round of verification, a comprehensive score is given according to the node’s communication ability and node consensus performance, and the transaction load in the shardings is calculated to confirm the transaction overload of the sharding, so as to enhance the processing capacity of the sharding in the next round of verification. Experiments show that the scheme can effectively deal with the transaction overload in the sharding, improve the transaction verification rate in the sharding, and improve the overall TPS(transaction per second), which provides a useful reference for the further research of the sharding.
    Reference | Related Articles | Metrics
    Hierarchical Lightweight Access Control Scheme in Cloud Environment
    TONG Qian, HE Heng, NIE Lei, ZHANG Panfeng
    Computer Engineering and Applications    2022, 58 (21): 109-118.   DOI: 10.3778/j.issn.1002-8331.2109-0202
    Abstract58)      PDF(pc) (779KB)(37)       Save
    Attribute-based encryption can realize fine-grained access control of ciphertext data and effectively solve the problem of data sharing in the cloud environment. Aiming at the problem that it is difficult for devices with limited computing capabilities to efficiently complete a large number of calculations in the attribute-based encryption process, this paper proposes a hierarchical lightweight access control scheme in cloud environment. This solution safely transfers most of the time-consuming encryption and decryption calculations to the cloud servers by introducing virtual attribute and dual keys, and optimizes the access structure. The data sharer only needs to encrypt multiple data with hierarchical access structure once, and the data requester can decrypt part or all of the ciphertext according to its attributes. Security analysis and performance evaluation show that the solution can achieve efficient and fine-grained ciphertext data access control in the cloud environment, significantly reducing the computing overhead of the client, and will not cause data leakage during the entire execution process.
    Reference | Related Articles | Metrics
    Research on Data Trading Mode Based on Ethereum and Dynamic Pricing
    CHEN Zhongwu, LING Jie
    Computer Engineering and Applications    2022, 58 (21): 119-130.   DOI: 10.3778/j.issn.1002-8331.2112-0465
    Abstract59)      PDF(pc) (1015KB)(30)       Save
    In order to solve the time-consuming and low efficiency of the transaction process in the existing data transaction model, information leakage and fair payment problems, an improved data transaction model is proposed. Additional constraints are preset through smart contracts, and data transactions and transactions are integrated. The function of arbitration dispute resolution is used to realize fair autonomy of transactions and control of transaction time to avoid malicious transactions in the process of data transactions. On this basis, in order to realize the dynamic balance of price in the proposed data trading model, based on the economic modeling method and the fairness and rationality of dynamic pricing, a dynamic pricing mechanism is designed to automatically balance aggregate supply and demand, and the price is dynamically adjusted according to the purchase demand and the market supply of data resources. The model is demonstrated from the dynamics of the model, which proves that the transaction price and demand can converge. This paper deploys and executes the contract based on the Ethereum experimental environment, and tests and analyzes the cost and security of each function of the smart contract. Simulation results show that the improved transaction mode can carry out data transaction with lower execution cost under dynamic pricing, and the smart contract has fewer code loopholes, which meets the feasibility and security requirements.
    Reference | Related Articles | Metrics
    Topological Scenario Emulation Technology of Space-Ground Integrated Satellite Network
    CHEN Xinyu, WANG Xiaofeng, LIU Yuan
    Computer Engineering and Applications    2022, 58 (20): 87-97.   DOI: 10.3778/j.issn.1002-8331.2108-0446
    Abstract98)      PDF(pc) (1098KB)(43)       Save
    Network emulation can provide strong support for the evaluation of new technologies of space-ground integrated satellite network. Aiming at the inherent heterogeneity and dynamic characteristics of the space-ground integrated satellite network topology scenario, a space-ground integrated satellite network topology scenario emulation technology is proposed. Firstly, a unified description model for heterogeneous and dynamic satellite network topologies is designed, and a method for automatic topology analysis and emulation scenario generation based on the unified description model is studied to improve the usability of emulation scenario generation. Secondly, from the perspective of time-varying characteristic, a method for classifying various inter-satellite and satellite-ground links is designed. The links are divided into time-varying and time-invariant. For time-invariant links, a link model preloading mechanism is introduced to improve the emulation efficiency and response speed of satellite links. Finally, in response to the low delay emulation accuracy in the process of link emulation, a link emulation remedy strategy is formulated to emulate real-time and dynamic satellite links with high fidelity. A variety of space-ground integrated satellite network scenarios are constructed. The experimental results indicate that the proposed technology has efficient automatic generation capability of satellite network emulation topologies and is obviously superior to existing technologies in the emulation efficiency and fidelity of the satellite links.
    Reference | Related Articles | Metrics
    Blockchain Traceability Scheme Based on Hierarchical Security QR Code
    ZHANG Xiaoyu, ZHANG Lina
    Computer Engineering and Applications    2022, 58 (20): 98-107.   DOI: 10.3778/j.issn.1002-8331.2110-0011
    Abstract100)      PDF(pc) (906KB)(46)       Save
    Food safety traceability system is of great significance to food safety and self-restraint of the food industry. Some existing traceability schemes that based on block chains have the problem of insufficient security for data storage and verification among communication nodes. Although some schemes use RFID to realize the access control between nodes, which can prevent information from being tampered with, the cost is too high and there is a lack of authentication between nodes. Some other schemes use QR code to record traceability information between nodes. Although such measures can reduce the cost, ordinary two-dimensional code cannot store confidential data. In this paper, a new traceability scheme based on block chain technology and secondary QR code is proposed for food safety traceability. A secure two-level QR code is designed to solve the security problem of data communication between nodes, and then a blockchain traceability scheme based on hierarchical secure QR code is proposed. In this scheme, the manager can verify the legitimacy of the member’s identity through the signature uploaded by members. Only legal members can upload QR code information. Other system participants can extract the group signature uploaded by the legal members to confirm the integrity of the information stored in the QR code. Compared with the traceability scheme using RFID and ordinary QR code, this paper realizes the functions of information tamper proof and identity authentication between nodes to solve the problem of information security in the process of node transaction.
    Reference | Related Articles | Metrics
    Anonymous Attribute?Based Encryption Scheme with Verifiable Outsourcing Decryption in Blockchain
    ZHUANG Chaoyuan, GUO Rui, YANG Geng
    Computer Engineering and Applications    2022, 58 (19): 124-134.   DOI: 10.3778/j.issn.1002-8331.2107-0487
    Abstract75)      PDF(pc) (1230KB)(37)       Save
    Ciphertext-policy attribute-based encryption(CP-ABE) is one of the key technology to achieve data confidentiality and fine-grained access control in a cloud environment. However, general CP-ABE schemes have some problems, such as the access policy sensitive information leakage, the high decryption cost and the too much power of attribute authority. In order to solve these problems, an anonymous attribute-based encryption scheme with verifiable outsourcing decryption based on blockchain(AVOCB-ABE) is proposed. The scheme uses policy-hiding idea to protect sensitive attribute information, uses both parties to generate the complete attribute key, and introduces attribute matching operation before decryption. In addition, the scheme uses the non-tampering feature of blockchain to storage verification parameters which can verify the outsourcing decryption results, the blockchain is used to generate and store attribute certificates. Finally, the security of selective ciphertext-policy and chosen-plaintext attacks is proved under the random oracle model, the scheme is compared with other schemes in function and communication overhead, simulated by using PBC Go cryptographic library, the results show that the scheme reduces the user’s decryption calculation.
    Reference | Related Articles | Metrics
    User Identity Authentication Model for Cross-Chain System
    WANG Sasa, DAI Bingrong, ZHU Menglu, LI Chao
    Computer Engineering and Applications    2022, 58 (19): 135-141.   DOI: 10.3778/j.issn.1002-8331.2107-0251
    Abstract83)      PDF(pc) (943KB)(33)       Save
    With the development and wide use of blockchain technology, the interconnection and interoperability between blockchains have become the focus of current research. Cross-chain identity authentication and management are the basis for trust interaction of blockchain. In order to solve the problems of security and performance in the current cross-chain identity authentication and management, a user identity and authentication model based on ECC-ZKP(elliptic curve cryptosystem-zero-knowledge proof) is proposed. By studying the cross-chain identity model, the elliptic curve encryption algorithm and zero-knowledge proof are introduced to realize the registration, update and authentication of cross-chain identity, and provide trusted identity authentication services for users’ cross-chain access and communication. The analysis and test show that the model has high security and good performance in processing efficiency and resource utilization.
    Reference | Related Articles | Metrics
    Group-Based Handover Authentication Scheme for 5G Heterogeneous Networks
    ZHANG Yinghui, LI Yiming, LI Yifei, ZHENG Dong
    Computer Engineering and Applications    2022, 58 (18): 137-146.   DOI: 10.3778/j.issn.1002-8331.2106-0495
    Abstract74)      PDF(pc) (1069KB)(62)       Save
    With the development of the fifth-generation mobile communication technology(5G), the quality of network services has been rapidly improving. However, the network environment is becoming more and more complex. Furthermore, it also brings more security challenges. The handover authentication can solve the problem of user access authentication between two different networks. But there are still some weaknesses in the existing schemes, such as universal handover authentication, key agreement, identity privacy, resistance to impersonation attacks, resistance to man-in-the-middle attacks, resistance to replay attacks and the group-based handover efficiency need to be improved. For these, a group-based handover authentication scheme for 5G heterogeneous networks is proposed. In the proposed scheme, the registered domain server stores a pass for each user on the blockchain. Using this data to authenticates the users can achieve the universal handover authentication. For group user access, each user sets aggregatable parameters separately. Then, the verifier performs batch verification by verifying the aggregate signature. By using the AVISPA tool to analyze the proposed protocol, it shows that the protocol is sufficiently secure. According to the performance analysis, the proposed scheme improves the efficiency by 89.8% compared with some existing schemes when performing batch verification.
    Reference | Related Articles | Metrics
    Directed Grey-Box Fuzzing Technology Based on LSTM and Dynamic Strategy
    LI Zhaoji, WANG Tianyuan, ZHOU Ziqiang, WANG Yao, CHEN Yongle
    Computer Engineering and Applications    2022, 58 (18): 147-153.   DOI: 10.3778/j.issn.1002-8331.2106-0507
    Abstract112)      PDF(pc) (658KB)(33)       Save
    Directed fuzzing is designed to quickly produce test cases, reach a series of given target locations, and discover program errors. However, the current directed fuzzing tools generally have the problem of low test efficiency. So a directed grey-box method based on neural network is proposed which builds a model to predict where the current seed can produce input gain by learning variation patterns in different locations in the input files from past fuzzing explorations, so as to guide the fuzzer to optimize mutation. At the same time, in order to solve the tradeoff of exploration-exploitation problem in directed fuzzers, a dynamic strategy is introduced to adaptively coordinate two stages in the process of fuzzy testing. A prototype system named DYNFuzz is implemented based on the existing fuzzing framework AFL, and is tested and evaluated on three benchmarks, which shows that DYNFuzz has higher directed performance and test efficiency than other fuzzers and would not be caught up in local dilemmas caused by the exploration-exploitation imbalance.
    Reference | Related Articles | Metrics
    Encrypted Malicious Traffic Detection Method Based on Transfer Learning
    ZHANG Surong, CHEN Bo, BU Youjun, LU Xiangyu, SUN Jia
    Computer Engineering and Applications    2022, 58 (17): 130-138.   DOI: 10.3778/j.issn.1002-8331.2106-0143
    Abstract70)      PDF(pc) (859KB)(45)       Save
    The existing encryption malicious traffic detection methods need to use a large number of accurately marked samples for training, to achieve a better detection effect. But in the real network environment, it is difficult to mark the encrypted traffic data correctly because its content is not visible. In view of the above problems, an encrypted malicious traffic detection method based on tranfer learning is proposed. The Eficientnet-B0, a pre-trained model based on the Imagenet dataset, is transferred to the encrypted traffic dataset for the first time. Its convolution layer structure and parameters are preserved, and the fully connected layers are replaced and retrained. By the idea of migration learning, the high detection performance under small sample condition is realized. Utilizing the end-to-end framework design, this method can extract the features from the original traffic data directly, then detect and classify them in fine-grained way, which avoids the complicated manual feature extraction process. The experimental results show that this method can achieve 99.87% binary classification accuracy and 98.88% fine-grained classification accuracy. Furthermore, when the number of various traffic samples in the training set is reduced to 100, it can also reach 96.35% of fine-grained classification accuracy.
    Reference | Related Articles | Metrics
    Fine-Grained Bilateral Access Control Scheme in Online Consultation Environment
    LI Yixin, ZHANG Yinghui, HU Lingyun, ZHENG Dong
    Computer Engineering and Applications    2022, 58 (17): 139-147.   DOI: 10.3778/j.issn.1002-8331.2106-0033
    Abstract83)      PDF(pc) (662KB)(41)       Save
    With the continuous development of online consultation technology, more and more patients choose to consult their conditions online. As patients usually consult their conditions twice or even several times online, this will not only lead to the leakage of medical record information, but also make the workload of medical staff increase dramatically. Therefore, it is necessary to encrypt the patient’s medical record information and improve the working efficiency of medical staff under this condition. At present, the existing ABE scheme can only protect their privacy information by selecting medical staff through the access control strategy developed by patients, and medical staff can only retrieve the information they need one by one from a large number of medical records, resulting in a sharp increase in their workload. To solve the above problems, a ciphertext policy attribute-based encryption scheme supporting fine-grained bilateral access control is proposed, and the data is stored by combining blockchain technology and IPFS storage technology. In this scheme, the patient’s medical record information is encrypted and uploaded to the IPFS system, and the unique hash index generated by the IPFS system is uploaded to the blockchain. Attribute-based encryption is used to protect user privacy and achieve fine-grained bilateral access control. The security analysis shows that the scheme is indistinguishable under the selective plaintext attack in the random oracle-machine model. Simulation results show that the proposed scheme improves the user’s computing efficiency compared with similar schemes.
    Reference | Related Articles | Metrics
    Unsupervised Link Prediction Algorithm Fusing Node Importance
    FU Xinyu, GU Yijun
    Computer Engineering and Applications    2022, 58 (16): 94-101.   DOI: 10.3778/j.issn.1002-8331.2105-0425
    Abstract113)      PDF(pc) (766KB)(47)       Save
    Improving the accuracy of weighted network link prediction algorithm is one of the basic problems in the study of complex networks. The common unsupervised prediction methods based on local network structure do not take into account that the nodes are more important, the they are more likely to generate new connections, and nodes with lower centrality are also highly important in real networks. To solve these problems, an unsupervised link prediction algorithm fusing node importance is proposed. In this algorithm, the possibility of new connections is calculated from the perspectives of structural similarity and node importance, and the influence degree is adjusted by using the custom coefficient. Experimental results on five real weighted network datasets show that, in solving the problem of fast prediction of small-scale weighted networks, the algorithm is more accurate than the same type of methods, and the supervised link prediction method is not applicable.
    Reference | Related Articles | Metrics
    Intrusion Detection Model Based on Improved Double Deep Q-Network
    WU Yali, WANG Junhu, ZHENG Shuailong
    Computer Engineering and Applications    2022, 58 (16): 102-110.   DOI: 10.3778/j.issn.1002-8331.2105-0402
    Abstract88)      PDF(pc) (783KB)(60)       Save
    As an effective defense method of network security, intrusion detection technology is an essential part of network security system. With the drastic development of the Internet, the amount of network data increases rapidly, and network attacks tend to be more complex and diversified, consequently, current intrusion detection technologies cannot identify all kinds of attacks effectively. Owing to the unbalanced problem between normal traffic and attack traffic in the real network environment and the low detection rate of attack traffic, this paper proposes a CBL_DDQN detection model based on improved double deep Q-network which is based on deep reinforcement learning. A hybrid model consisting of one-dimensional convolutional neural network and bi-directional long short-term memory network is utilized in the DDQN framework of deep reinforcement learning, then the feedback learning and strategy-generating mechanism of deep reinforcement learning is used for training the agent to classify different types of attack samples, which can greatly weaken the dependence on data labels in the process of training model. In the meantime, the Borderline-SMOTE algorithm is used to reduce data imbalance so as to improve the detection rate of rare attack traffic. The performance of the model evaluated by NSL_KDD and UNSW_NB15 datasets shows that the model performs well in accuracy, precision and recall. The detection result of the model is far better than that of Adam-BNDNN, KNN, SVM and other detection methods, which implies the intrusion detection model proposed in this paper is efficient.
    Reference | Related Articles | Metrics
    Bitcoin Network Abnormal Transaction Detection Based on LSCP Algorithm
    LIAO Qian, GU Yijun
    Computer Engineering and Applications    2022, 58 (15): 117-123.   DOI: 10.3778/j.issn.1002-8331.2105-0028
    Abstract85)      PDF(pc) (682KB)(35)       Save
    Anomaly detection is one of the research hotspots in Bitcoin transaction data analysis. In view of the problems that the existing abnormal transaction detection methods based on machine learning are difficult to accurately summarize various abnormal types, and the generalization ability is insufficient, the network structure of Bitcoin transaction data is constructed and the related features of abnormal behavior patterns are extracted, and the detection model is constructed by using the parallel integration algorithm(LSCP) based on local dynamic selection and combination, and seven classical abnormal detection algorithms are incorporated into the algorithm, so as to improve the reliability and stability of the detection model by using the sensitivity of the base learner to different abnormal types. The experimental results show that, compared with the traditional detection method, the LSCP algorithm combined with the heterogeneous learner has a better effect on the overall detection performance.
    Reference | Related Articles | Metrics
    MTD Enhanced Cyber Deception Defense System
    GAO Chungang, WANG Yongjie, XIONG Xinli
    Computer Engineering and Applications    2022, 58 (15): 124-132.   DOI: 10.3778/j.issn.1002-8331.2105-0169
    Abstract103)      PDF(pc) (746KB)(28)       Save
    Computer networks are developing rapidly, but network security such as system damage and information leakage are also becoming increasingly prominent. Attackers usually conduct a large number of network reconnaissance before a formal attack to discover exploitable vulnerabilities in the target network and system. The static configuration in traditional network systems provides a great advantage for adversaries to find network targets and launch attacks. To reduce the effectiveness of adversaries’ continuous reconnaissance attacks, this paper develops a moving target defense enhanced cyber deception defense system based on software-defined networks. The system uses cyber deception technology to confuse the target network and system information collected by the attacker, extends the time for the attacker to scan the real vulnerable hosts in the network, and increases the attacker’s time cost. Besides, this paper integrates IP address randomization technology on the cyber deception, dynamically and randomly changes the IP addresses of nodes in the network to enhance the defensive effectiveness of the network deception system. Finally, the system prototype is implemented and evaluated. In a configuration where the virtual network topology scale is three network segments, and the address conversion cycle is 30 seconds, this system delays the adversaries’ discovery of vulnerable hosts by an average of seven times, reducing the probability of adversaries successfully attacking vulnerable hosts by 83%. At the same time, the system overhead is less than 8% on average.
    Reference | Related Articles | Metrics
    Application of Prophet Mixture Model on Long-Term Prediction of Base Station Cell Network Traffic
    ZHANG Jiachen, ZUO Xingquan, HUANG Hai, HAN Jing, ZHANG Baisheng
    Computer Engineering and Applications    2022, 58 (14): 80-88.   DOI: 10.3778/j.issn.1002-8331.2102-0111
    Abstract98)      PDF(pc) (1553KB)(64)       Save
    Most of the traditional network traffic prediction methods focus on short-term, while long-term prediction can better guide the base station cell wireless equipment expansion and contraction. Ensemble empirical mode decomposition(EEMD) can convert the non-stationary time series into a stationary time series, and Prophet model can accurately make a more accurate long-term prediction of traffic series. Based on the advantages of the above model methods and the non-linear and non-stationary characteristics of the base station cell network traffic, Prophet hybrid EEMD method(E-Prophet) is proposed for base station cell network traffic prediction. First, the EEMD is adopted to decompose the network traffic series into several intrinsic mode functions(IMF) components and a residual component; then, the Prophet model is used to model each component, and the predictions of each component are linearly combined to obtain the final prediction result. The method is verified using real-world base station cell network traffic data, and the results show that E-Prophet model has higher accuracy and robustness in long-term prediction of network traffic compared with Prophet, SARIMA, LSTM, and the model combining EMD and Prophet.
    Reference | Related Articles | Metrics
    Malicious Traffic Detection Method Based on Siamese Neural Network
    LI Daoquan, LU Xiaofu, YANG Qianqian
    Computer Engineering and Applications    2022, 58 (14): 89-95.   DOI: 10.3778/j.issn.1002-8331.2104-0159
    Abstract102)      PDF(pc) (750KB)(51)       Save
    With the development of technology, PC and cell phones have become indispensable smart devices in modern society. A large number of applications in PC and cell phones provide users with convenient web services such as live chat, email, and downloads, etc. via the Internet. However, the popularity of these devices has also attracted a large number of malicious attackers, and malicious applications and malicious traffic have arisen as a result. To address this problem, this paper proposes an end-to-end single-sample detection method based on Siamese neural network on the basis of malicious traffic classification detection. First, the sample data is pre-processed into grayscale images, then the image samples are trained and learned in the TensorFlow deep learning framework, and finally the detection of malicious traffic is achieved by comparing the similarity between grayscale images. The method proposed in this paper can not only realize end-to-end single-sample detection, but also provides a solution to the classification problem of imbalanced dataset. The final experimental detection accuracy rate can reach 95.04%, which proves the feasibility and scientific validity of this method.
    Reference | Related Articles | Metrics
    Efficient and Revocable Fog-Assisted Cloud Access Control Scheme
    SUN Xiao, WANG Zheng, LI Ling
    Computer Engineering and Applications    2022, 58 (13): 112-118.   DOI: 10.3778/j.issn.1002-8331.2102-0181
    Abstract88)      PDF(pc) (739KB)(36)       Save
    Ciphertext-policy attribute-based encryption not only realizes the fine-grained access control of data in IoT system based on cloud storage, but also brings the problem of user and attribute revocation. However, in the existing access control schemes, the time-based schemes are difficult to achieve immediate revocation, and the third-party-based schemes usually require a large number of re-encrypted ciphertexts, the efficiency is low and the cost is large. Therefore, an efficient access control scheme supports immediate revocation of user and attribute based on RSA key management mechanism is proposed. The length of the keys and ciphertexts are fixed. With the help of fog nodes, user revocation is realized. At the same time, part of the encryption and decryption work is unloaded from the client to the nearby fog node, which reduces the computing burden of the client. The results of security analysis based on aMSE-DDH hypothesis show that the scheme can resist chosen-ciphertext attack. Theoretical analysis and experiments prove that the proposed scheme can provide efficient access control for application scenarios with frequent user and attribute changes and limited resources.
    Reference | Related Articles | Metrics
    Intrusion Detection Algorithm Combining Convolutional Neural Network and Three-Branch Decision
    WU Qirui, HUANG Shucheng
    Computer Engineering and Applications    2022, 58 (13): 119-127.   DOI: 10.3778/j.issn.1002-8331.2103-0443
    Abstract84)      PDF(pc) (737KB)(52)       Save
    With the diversification and intelligentization of network intrusion behaviors, traditional intrusion detection algorithms have problems such as inadequate feature extraction and inaccurate model classification when they process massive data with high-dimensional feature and non-linearity. Therefore, an intrusion detection algorithm combining convolutional neural network(CNN) and three-way decision(TWD) is proposed. Convolutional neural network has superior feature extraction ability. At the same time, three-way decision can avoid the risk caused by blind classification due to insufficient information, and reduce the time in classification. This method uses the convolutional neural network to extract high-dimensional data feature and constructs multi-granularity feature space. Then, the real-time decision on network behavior will be made through the theory of three-way decision. When the network behavior cannot be decided immediately, deferred decisions are made. In other words, this part of network behavior will be extracted again to construct feature spaces of different granularity. Finally, the classification results will be output. Experimental results on NSL-KDD and CIC-IDS2017 data sets show that the proposed algorithm can improve the performance of intrusion detection system.
    Reference | Related Articles | Metrics
    Codeword Replacement Based on Shortest Euclidean Distance for VoIP Steganography
    SUN Xinhao, WANG Kaixi
    Computer Engineering and Applications    2022, 58 (13): 128-134.   DOI: 10.3778/j.issn.1002-8331.2102-0291
    Abstract108)      PDF(pc) (892KB)(24)       Save
    Imperceptibility is the primary goal of information hiding technology, reducing the modification rate is an effective way to improve imperceptibility. The codeword replacement based on shortest Euclidean distance for VoIP steganography is proposed based on the principle that the less modification to a cover will achieve the high imperceptibility, which decreases the modification to a cover and improves the steganography efficiency, thus the invisibility of steganography is enhanced without reducing the quality. After a steganography unit is defined, the Euclidean distance between the secret message and the linear predictive coding(LPC) parameters in the unit is calculated, and the LPC parameter with the shortest distance is selected and replaced by the secret message. The complementary neighbor vertex(CNV) method is employed to indicate which LPC parameters are modified. Compared with other methods, the experimental results show that the proposed method has higher steganography efficiency, its steganography modification rate reduce to 22.7%, its KL divergence indicates the better security, and its PESQ exceeds 4.0, which shows good speech quality.
    Reference | Related Articles | Metrics
    Image Steganography Location Research Based on Pixel Probability Prediction
    CHEN Sheng, LI Zhi
    Computer Engineering and Applications    2022, 58 (12): 85-93.   DOI: 10.3778/j.issn.1002-8331.2101-0422
    Abstract125)      PDF(pc) (1308KB)(34)       Save
    In order to further enhance the practicability of image steganalysis, this paper expands the research goal of image steganalysis as steganography location of adaptive steganography and non-adaptive steganography LSB matching, an end-to-end steganography localization network PSL_NET is proposed. Input an image at the input end, and locate the position of the steganographic pixel of the image at the output end. In the preprocessing layer, the high-pass filter of the spatial rich model is used to extract the residual noise images. In the depth residual layer, deep residual learning is used to enhance the expression ability of steganographic features. In the pixel prediction layer, using the mask image which marks the actual position of the steganographic pixel to perform supervised learning, as well as treating the pixels in the smooth or texture area without distinction, the probability whether each pixel is steganographic pixels is predicted, and finally predict the steganographic pixels of the input image. The imbalance problem of positive and negative samples are solved from the perspective of objective function to improve the detection accuracy. In the experiment based on BOSSbase v1.01, when the network predicts the steganographic image of the adaptive steganography algorithm S-UNIWARD at the payload of 0.4 BPP, the pixel detection accuracy is 0.981 74, and the experiments also verify the network can detect the steganographic image embedded by non-content adaptive steganography LSB matching.
    Reference | Related Articles | Metrics
    Representative User Sampling Algorithm Based on Weighted Neighborhood
    HE Shuimiao, BAN Zhijie
    Computer Engineering and Applications    2022, 58 (12): 94-101.   DOI: 10.3778/j.issn.1002-8331.2012-0554
    Abstract81)      PDF(pc) (1053KB)(36)       Save
    The representative user sampling method is widely used in the field of social network analysis, how to make its subset represent all users in the network has great significance. The existing methods pay less attention to the large amount of useful information of potential users in the network topology, by optimizing the statistical stratified sampling model, the representative user sampling algorithm based on weighted neighborhood is proposed. In order to get more valuable content from the network topology, the algorithm uses the weighted neighborhood to improve the calculation method of user representation, and combines with user attributes. Then users are divided into different attribute groups according to their attribute values, and the representation of users in each attribute group is calculated. After that, the representation of representative users is measured by quality function. The heuristic greedy algorithm is used to extract representative users. By comparing with six traditional sampling algorithms on four data sets, the results show that the representative user sampling algorithm based on weighted neighborhood improves the accuracy rate, recall rate and F1-Measure evaluation index.
    Reference | Related Articles | Metrics