����Ŀ¼

15 April 2022, Volume 8 Issue 2

Previous Issue    Next Issue

Comprehensive Review

Survey on intellectual property protection for deep learning model

Xinya WANG, Guang HUA, Hao JIANG, Haijian ZHANG

2022, 8(2):  1-14.  doi:10.11959/j.issn.2096-109x.2022015

Asbtract ( 1248 )   HTML ( 185)   PDF (1274KB) ( 1328 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

With the rapid development of deep learning technology, deep learning models have been widely used in many fields such as image classification and speech recognition.Training a deep learning model relies on a large amount of data and computing power, thus selling the trained model or providing specific services (DLaaS, e.g.) has become a new business.However, the commercial interests of model trainers and the intellectual property rights of model developers may be violated if the model is maliciously stolen.With deep neural network watermarking becoming a new research topic, multimedia copyright protection techniques were used for deep learning model protection.Numerous methods have been proposed in this field and then a comprehensive survey is needed.the existing deep neural network watermarking methods were elaborated and summarized and the future research directions of this field were discussed.The overall framework of neural network watermarking was presented, whereby the basic concepts such as classification model and model backdoor were introduced.Secondly, the existing methods were divided into two types according to the mechanism of watermark embedding, one is to embed the watermark bits into the carrier of internal information of the network, and the other one uses the established backdoor mapping as the watermark.These two existing deep neural network watermarking methods were analyzed and summarized, and attacks to the watermarks were also introduced and discussed.By analyzing the white-box and black-box conditions in watermarking scenario, it comes to the conclusion that the model is difficult to be effectively protected when it is distributed in the white-box manner, and the neural network watermark defenses in the black-box distribution and black-box verification are both worthy for further research.

Topic: Cybersecurity——Attack and Defense Technologies

Survey on the development of mimic defense in cyberspace:from mimic concept to “mimic+” ecology

Hailong MA, Liang WANG, Tao HU, Yiming JIANG, Yanze QU

2022, 8(2):  15-38.  doi:10.11959/j.issn.2096-109x.2022018

Asbtract ( 627 )   HTML ( 98)   PDF (2454KB) ( 693 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Build upon the dynamic-heterogeneous-redundant architecture for multi-body execution, multi-mode ruling and multi-dimension reconstruction, cyberspace mimic defense (CMD) uses uncertain system to deal with the uncertain threat to cyberspace ubiquity.The evolution of CMD over the past 8 years were reviewed systematically from the vertical, horizontal, current, developing and future perspectives.From the vertical perspective, the development process of CMD from concept to theory and practice was summarized.From the horizontal view, it elaborated the core structure DHR (dynamical heterogeneous redundancy) of CMD, the principles based on CMD’s three major theorems, safety gains and performance costs.A comparison was conducted between CMD and five other active defense technologies, namely intrusion tolerance, moving target defense, zero trust architecture, trusted computing and computer immunology.From the current perspective, it reviewed the implementation elements, performance, system architecture, heterogeneous strategies, scheduling strategies, voting strategies and other common technology patterns and characteristics of 11 types of main existing mimic products including mimicry router, mimicry processor, mimicry DNS server and mimicry cloud platform.From the developing perspective, it explored the “mimic+” AICDS (Mimic + AI/IoT/Cloud/Data/SDN) symbiotic ecology with respect to 5 types of new technologies, namely artificial intelligence, Internet of things, cloud computing, big data and software-defined network, and proposed the corresponding technology junctions and cross research value.From the future perspective, it looked into the future mimicry baseline 2.0 product ecology, “mimic +5G/6G”, “mimic + edge computing”,“mimic + cloud” and “mimic + blockchain” application scenarios.Besides, 4 types of challenges faced by CMD in the future were analyzed and summarized, including escape space of multi-mode decision attack, mutual restriction of heterogeneous and synchronous, difficult balance between security and function, and limited transformation space of existing endogenous security components.

Research on IoT security situation awareness method based on evidence theory

Jian LI, Tinglu DONG, Jie LI

2022, 8(2):  39-47.  doi:10.11959/j.issn.2096-109x.2022022

Asbtract ( 306 )   HTML ( 61)   PDF (1640KB) ( 222 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The security problem of IoT became more and more serious with its rapid development.Considering that the current IoT security situation awareness system lacks generality and excessively relies on expert knowledge, a IoT security situation awareness method based on improved D-S evidence theory was proposed in this paper.Fuzzy Gaussian membership function was used to calculate the vulnerability information membership matrix, which was normalized as evidence distribution matrix.The improved Topsis method was used to measure the evidence credibility.In order to fully restrain the credibility of conflicting evidence and improve the credibility of mutually supporting evidence, local credibility between two evidence was aggregated and the expected positive and negative solution vectors were improved according to the situation assessment scenario.And the weighted average method was used for vulnerability information fusion, to obtain the result of situational assessment.The result of situational awareness was fused with the time discount and high-risk vulnerability information discount evidence theory.At the same time, the IoT vulnerability information at different moments was considered comprehensively, the evidence was adaptively and dynamically weighted with the ratio information of high-risk vulnerability.The experimental results show that in the fusion of different numbers of evidence bodies and four common conflicting evidence, the improved Topsis method has higher fusion probability on credible proposition.In the aspect of situation assessment, the risk degree of current system is accurately assessed.And in the aspect of situational awareness, this discount evidence theory can predict the probability of high risk and critical risk, which is more effective than the traditional D-S evidence theory.According to this theory, a IoT security situational awareness method process was proposed, which would be used to guide engineering practice.In the future, the relationship between vulnerabilities can be considered and richer information between vulnerabilities can be extracted for vulnerability exploiting, so that the result of situation assessment is more accurate and reasonable.On the other hand, for situational awareness, game theory can be adopted in the process of dynamic game between the attacker and defender.

Novel defense based on softmax activation transformation

Jinyin CHEN, Changan WU, Haibin ZHENG

2022, 8(2):  48-63.  doi:10.11959/j.issn.2096-109x.2022016

Asbtract ( 308 )   HTML ( 44)   PDF (5759KB) ( 258 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Deep learning is widely used in various fields such as image processing, natural language processing, network mining and so on.However, it is vulnerable to malicious adversarial attacks and many defensive methods have been proposed accordingly.Most defense methods are attack-dependent and require defenders to generate massive adversarial examples in advance.The defense cost is high and it is difficult to resist black-box attacks.Some of these defenses even affect the recognition of normal examples.In addition, the current defense methods are mostly empirical, without certifiable theoretical support.Softmax activation transformation (SAT) was proposed in this paper, which was a light-weight and fast defense scheme against black-box attacks.SAT reactivates the output probability of the target model in the testing phase, and then it guarantees privacy of the probability information.As an attack-free defense, SAT not only avoids the burden of generating massive adversarial examples, but also realizes the advance defense of attacks.The activation of SAT is monotonic, so it will not affect the recognition of normal examples.During the activation process, a variable privacy protection transformation coefficient was designed to achieve dynamic defense.Above all, SAT is a certifiable defense that can derive the effectiveness and reliability of its defense based on softmax activation transformation.To evaluate the effectiveness of SAT, defense experiments against 9 attacks on MNIST, CIFAR10 and ImageNet datasets were conducted, and the average attack success rate was reduced from 87.06% to 5.94%.

Defense scheme for the world state based attack in Ethereum

Zhen GAO, Dongbin ZHANG, Xiao TIAN

2022, 8(2):  64-72.  doi:10.11959/j.issn.2096-109x.2022013

Asbtract ( 281 )   HTML ( 40)   PDF (1112KB) ( 378 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Ethereum is taken as the representative platform of the second generation of blockchain system.Ethereum can support development of different distributed applications by running smart contracts.Local database is used to store the account state (named world state) for efficient validation of transactions, and the state root is stored in the block header to guarantee the integrity of the state.However, some researches revealed that the local database could be easily tempered with, and attackers can issue illegal transactions based on the modified account state to obtain illegitimate benefits.This world-state based security problem was introduced, and the preconditions for attack were analyzed.Compared with the two common security threats under the PoW (proof of work) consensus, it was found that when the attacker controls the same mining computing power, the world-state based attack brought higher risk, and the success rate approached 100%.In order to deal with this threat, a practical scheme for attack detection and defense was proposed accordingly.The secondary verification and data recovery process were added to the Ethereum source code.The feasibility and complexity of the proposed scheme was evaluated with single-machine multi-threading experiments.The proposed scheme improves Ethereum’s tolerance to malicious tampering of account state, and is applicable to other blockchain platforms applying local database for transaction validation, such as Hyperledger Fabric.In addition, the time and computational overhead brought by the proposed scheme are not prominent, so it has good applicability and induces acceptable impact on the performance of original system.

Defense mechanism of SDN application layer against DDoS attack based on API call management

Yang WANG, Guangming TANG, Shuo WANG, Jiang CHU

2022, 8(2):  73-87.  doi:10.11959/j.issn.2096-109x.2022017

Asbtract ( 296 )   HTML ( 43)   PDF (2854KB) ( 445 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Due to thelack of strict access control, identity authentication and abnormal call detection, attackers may develop malicious applications easily and then it leads to theabuse of the northbound interface API (application programming interface) accordingly.There are mainly two patterns of DDoS (distributed denial-of-service) attacks against application layer.1) malicious App bypass the security review of the northbound interface and make a large number of calls to some API in a short time, thus causing the controller to crash and paralyzing the whole network; 2) attackers take a legitimate SDN (software defined network) application as the target and make a large number of short-time calls to the specific API needed by the application, which makes the legitimate App unable to call the API normally.Compared with the first pattern, the second one is more subtle.Therefore, it’s necessary to distinguish whether the App is malicious or not, effectively clean the App running on the attacked controller, and redistribute the controller to the legitimate App.Based on the in-depth analysis of the development trend of the current northbound interface, the possible DDoS attack patterns were simulated and practiced.Then a DDoS defense mechanism for SDN application layer was proposed.This mechanism added an App management layer between SDN application layer and control layer.Through reputation management, initial review, mapping allocation, anomaly detection and identification migration of the App, the malicious App attack on SDN can be predicted and resisted.The proposal focused on pre-examination of malicious App before attacks occur, so as to avoid attacks.If the attack has already happened, the operation of cleaning and separating the legitimate App from the malicious App is triggered.Theoretical and experimental results show that the proposed mechanism can effectively avoid DDoS attacks in SDN application layer, and the algorithm runs efficiently.

Adversarial examples defense method based on multi-dimensional feature maps knowledge distillation

Baolin QIU, Ping YI

2022, 8(2):  88-99.  doi:10.11959/j.issn.2096-109x.2022012

Asbtract ( 321 )   HTML ( 36)   PDF (3603KB) ( 376 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The neural network approach has been commonly used in computer vision tasks.However, adversarial examples are able to make a neural network generate a false prediction.Adversarial training has been shown to be an effective approach to defend against the impact of adversarial examples.Nevertheless, it requires high computing power and long training time thus limiting its application scenarios.An adversarial examples defense method based on knowledge distillation was proposed, reusing the defense experience from the large datasets to new classification tasks.During distillation, teacher model has the same structure as student model and the feature map vector was used to transfer experience, and clean samples were used for training.Multi-dimensional feature maps were utilized to enhance the semantic information.Furthermore, an attention mechanism based on feature map was proposed, which boosted the effect of distillation by assigning weights to features according to their importance.Experiments were conducted over cifar100 and cifar10 open-source dataset.And various white-box attack algorithms such as FGSM (fast gradient sign method), PGD (project gradient descent) and C＆amp;W (Carlini-Wagner attack) were applied to test the experimental results.The accuracy of the proposed method on Cifar10 clean samples exceeds that of adversarial training and is close to the accuracy of the model trained on clean samples.Under the PGD attack of L2 distance, the efficiency of the proposed method is close to that of adversarial training, which is significantly higher than that of normal training.Moreover, the proposed method is a light-weight adversarial defense method with low learning cost.The computing power requirement is far less than that of adversarial training even if optimization schemes such as attention mechanism and multi-dimensional feature map are added.Knowledge distillation can learn the decision-making experience of normal samples and extract robust features as a neural network learning scheme.It uses a small amount of data to generate accurate and robust models, improves generalization, and reduces the cost of adversarial training.

Papers

Universal patching method for side-channel vulnerabilities based on atomic obfuscation

Deqing ZOU, Pan ZHANG, Wei LIU, Weijie CHEN, Yifan LU

2022, 8(2):  100-111.  doi:10.11959/j.issn.2096-109x.2022014

Asbtract ( 235 )   HTML ( 24)   PDF (1202KB) ( 399 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Executing code containing side-channel vulnerabilities exhibits different non-functional behaviors related to inputs.Attackers can obtain these behaviors by leveraging micro architecture side-channel attacks and then analyze the pattern between the behaviors and the inputs to access sensitive data.Vulnerability repairing at the software layer brings low overheads to a program’s execution.Besides, it does not require modifying hardware or system, which enables fast patching and widespread deployment.It becomes the mainstream strategy applied to the current cryptographic implementations.However, existing solutions are deeply bound to the program’s implementation and requires manual intervention.This brings challenge to implement and is not versatile enough.A general patching method was proposed for side-channel vulnerabilities that combined dynamic obfuscated execution with hardware atomic transaction.To hide the real accesses of the side-channel vulnerabilities of a program, the proposed method inserted dynamic confusing accesses into the vulnerabilities.To avoid an attacker using fine-grained side-channel attack to distinguish the real access and the confusing access, both of them were encapsulated as transactions and they were guaranteed to be uninterrupted during the running period.In addition, a prototype system called SC-Patcher was implemented based on the LLVM compiler.Various optimization strategies were supported, including secure springboard and transaction aggregation, to further improve system security and performance.Experimental results show that the proposed method makes it impossible for an attacker to restore accurate sensitive data through side-channel attack, and it also brings almost no additional performance overhead to the program.

Searchable encryption scheme based on attribute policy hiding in a cloud environment

Yihua ZHOU, Xinyu HU, Meiqi LI, Yuguang YANG

2022, 8(2):  112-121.  doi:10.11959/j.issn.2096-109x.2022019

Asbtract ( 312 )   HTML ( 48)   PDF (1380KB) ( 462 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Attribute-based searchable encryption technology can achieve fine-grained access control of data, but the existing searchable encryption scheme, keyword search, access control and file encryption are basically performed separately, causing the attacker to directly skip the access policy for keyword index matching and file decryption.Besides, the data owners in the existing schemes need to pass the key of the encrypted file to the user in a secure channel, which increases the cost of the data owner.Furthermore, most tree-based access control policies are open and easy to cause privacy leakage.Therefore, based on the LSSS (linear secret sharing schemes) access architecture, the searchable encryption scheme based on attribute policy hiding in a cloud environment was proposed.Through the embedding of policy secret values into keyword encryption and file storage encryption, the combination of access control, keyword search and file encryption were realized.The aggregate key technology enables users to decrypt files without interacting with the data owner, reducing the burden of key management and increasing storage space by approximately 30%.The experimental results and security analysis show that the proposed scheme guarantees the security of stored data, privacy of access strategy and non-connectivity of trap gate.Compared with the existing mainstream scheme, the retrieval efficiency of the proposed scheme has improved to more than 20%.

Electronic invoice public verification scheme based on SM2 coalition signature algorithm

Yurong LUO, Jin CAO, Hui LI, Xingwen ZHAO, Chao SHANG

2022, 8(2):  122-131.  doi:10.11959/j.issn.2096-109x.2022020

Asbtract ( 407 )   HTML ( 38)   PDF (2939KB) ( 201 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

In order to solve the current problems of electronic invoices, such as anti-counterfeiting, privacy leakage and low verification efficiency, an electronic invoice public verification architecture based on the national signature algorithm was proposed.Electronic invoice documents have complex data sources, sensitive user information on the ticket surface, frequent data flow and other characteristics as well as the demand for efficient public verification.Based on these motivations, the e-invoice generation and verification protocol was designed.Besides, a signature code generation scheme for invoice anti-counterfeiting was proposed based on certificate-free joint signature.Then multiple data verification and signatures of the invoicing party and the taxation authority can be realized.All kinds of invoice holders can not only verify the authenticity and data integrity of electronic invoices, but also publicly check the authenticity and data integrity of electronic invoices.The integration of data encryption algorithms protects user privacy data in electronic invoices, and at the same time can fulfill the requirements for authenticity or status verification under various invoice application scenarios.The checking party only needs to verify one signature to confirm the authenticity of the electronic invoice signed by both parties.Scyther security simulation tools were used to analyze the security of the proposed solution, which can verify the integrity and authenticity of the data source and check the confidentiality of the privacy data under various types of attacks.Compared with a typical overseas e-invoice checking scheme and a similar digital signature-based e-invoice checking scheme, the proposed scheme has obvious advantages in terms of checking efficiency and invoice file size.

Research on evolution and security risk of Metaverse

Hailong WANG, Yangchun LI, Yuxiao LI

2022, 8(2):  132-138.  doi:10.11959/j.issn.2096-109x.2022023

Asbtract ( 798 )   HTML ( 95)   PDF (634KB) ( 710 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Recently, the Metaverse boom has swept the world.To improve the voice in the future digital field, major countries and well-known enterprises have accelerated arrangement of this emerging industry.However, there is no unified definition and ultimate form description of Metaverse.It is urgent to understand the essential characteristics, evolution trend and security risks of Metaverse scientifically, which can promote the high-quality development of China’s electronic information industry and ensure the cyberspace security.With literature research and related information analysis, the definition of Metaverse was put forward based on the core concepts such as technology and application iteration, economic and business model and the role of social relations.Additionally, the development and evolution of Metaverse were divided into four stages (data creation, digital bionics, virtual mirror and virtual-real symbiosis) according to the evolution degree of space category, space-time dimension, key technology and virtual real interaction.And the characteristics and corresponding forms of each stage were expounded.Then it analyzed the main reasons for the popularity of the Metaverse from the perspectives of innovative application demand, virtual life transformation, commercial value driving and political power driving.Focusing on the problems of challenging the national governance system, impacting the mainstream ideology, manipulating the capital market, causing potential social problems and misleading the layout of technology industry, it analyzed the multiple security risks and threats brought by the Metaverse wave from perspectives including politics, culture, economy, society and technology.It is suggested that the Metaverse is still in the stage of concept improvement and product exploration now.China should rationally grasp the development opportunities of the Metaverse based on the current situation of the electronic information industry, prepare for risks in advance, build a digital governance supervision system, stimulate industrial innovation and open development, consolidate the support of network information security, and create a healthy and orderly international environment.

Two-stage community detection algorithm based on label propagation

Xueliang SUN, Wei WANG, Junheng HUANG, Guodong XIN, Bailing WANG

2022, 8(2):  139-149.  doi:10.11959/j.issn.2096-109x.2021099

Asbtract ( 311 )   HTML ( 32)   PDF (1459KB) ( 194 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Community detection is an important research topic of complex network analysis.The detection results help to understand the community structure of complex networks and provide support for downstream tasks, such as content recommendation, link detection.Considering the challenge of community detection in complex networks, a two-stage community detection algorithm based on label propagation (TS-LPA) was proposed.The TS-LPA algorithm quantified the propagation capability of nodes based on the extended neighborhood.Then a new evaluation index was proposed to measure the probability of influence between nodes using the information of nodes and the weight of edges in the network.Based on the calculation of node centrality, the algorithm determined the updating sequence of node labels and the selection strategy of seed nodes, which eliminated the instability of the algorithm in the updating process.The TS-LPA algorithm used the breadth-first propagation idea and introduced the second-stage label propagation method to improve the quality of community detection.When label began to spread, all neighboring nodes had an influence on the label of the related node.Meanwhile, the influence of neighboring seed nodes was added to complete label updating, in order to reduce the dominance of neighboring nodes on the updated node.The experimental results of different real data sets and synthetic data sets show that TS-LPA algorithm can eliminate randomness and show strong stability while effectively improving the quality of community detection.

Evidence classification method of chat text based on DSR and BGRU model

Yu ZHANG, Binglong LI, Xuejuan LI, Heyu ZHANG

2022, 8(2):  150-159.  doi:10.11959/j.issn.2096-109x.2022007

Asbtract ( 367 )   HTML ( 15)   PDF (1752KB) ( 347 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

It is always unlikely to efficiently identify and extract chat text evidence related to criminal events, due to the complex semantics such as “slang” in the chat content and the huge amount of chat text data generated by social software such as instant messaging.Based on this motivation, a chat text evidence classification model (DSR-BGRU) based on the DSR (dynamic semantic representation) model and the BGRU (bidirectional gated recurrent unit) model was proposed.The chat text data was pre-processed to preserve the characteristics of the criminal field.Then a multi-layer chat text feature extraction and classification model using the Keras framework was proposed.With the text matrix composed of vector representation of words in the DSR model as the input vector, the input layer of the DSR model featured the chat text from the semantic level.Then the hidden layer of the BGRU model extracted the context characteristics of the text composed of the word vectors.The softmax classification layer recognized and extracted the chat text evidence.The experimental results show that the proposed DSR-BGRU can more accurately identify and extract chat records compared with other models and methods for text classification, and it can also effectively extract the criminal text information from the chat information with the accuracy rate 92.06% and the F1 score 91.00%.

Java deserialization gadget chain discovery method based on hybrid analysis

Yongxing WU, Libo CHEN, Kaida JIANG

2022, 8(2):  160-174.  doi:10.11959/j.issn.2096-109x.2022009

Asbtract ( 985 )   HTML ( 121)   PDF (3513KB) ( 1041 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Java deserialization vulnerabilities have become a common threat to Java application security nowadays.Finding out the gadget chain determines whether this type of vulnerability can be exploited.Due to the large code space of Java applications and dependent libraries and the polymorphism of Java itself, manual analysis of Java deserialization gadget chains consumes a lot of time and effort and it is highly dependent on the experienced knowledge.Therefore, it is crucial to study how to efficiently and accurately automate the discovery of gadget chains.Java deserialization gadget chain discovery method based on hybrid analysis was proposed.Call graph based on the variable declaration type was constructed, and then the deserialization entry functions that may reach the dangerous functions were screened using the call graph analysis.The screened entry functions were used as the entry point of the hybrid information flow analysis.The hybrid information flow analysis was carried out for both pointer and tainted variables.The tainted objects created implicitly were marked.The tainted information and the pointer information were propagated simultaneously to construct the hybrid information flow graph.The reachability of external taint data propagation to the dangerous function was judged based on the hybrid information flow graph.The corresponding deserialization gadget chain was constructed according to the taint propagation path.The hybrid analysis took into account the efficiency of call graph analysis and the accuracy of hybrid information flow analysis.The corresponding static analysis tool, namely GadgetSearch, was implemented based on the proposed hybrid analysis method.In the experimental evaluation, GadgetSearch had lower false positive and lower false negative than the existing tool GadgetInspector on four datasets of Ysoserial, Marshalsec, Jackson historical CVE, and XStream historical CVE.Additionally, GadgetSearch also found multiple undisclosed gadget chains.The experimental results show that the proposed method can efficiently and accurately discover the Java deserialization gadget chain in multiple practical Java applications.

Trust evaluation optimization mechanism for cloud user behavior based on FANP

Yi ZHANG, Liqin TIAN, Zenan WU, Wenxing WU

2022, 8(2):  175-182.  doi:10.11959/j.issn.2096-109x.2021100

Asbtract ( 255 )   HTML ( 29)   PDF (888KB) ( 382 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

The open cloud computing environment is facing security challenges and the traditional user behavior evaluation mechanism cannot guarantee the security of the cloud.In order to scientifically and quantitatively evaluate the user’s behavior trust, ensure the scientific and reasonable weight assignment, improve the security and credibility of user behavior under the cloud platform, a trust evaluation optimization mechanism combined with fuzzy analytic network process (FANP) was designed.In the proposal, user behavior trust evaluation based on one control target was extended to include two control target modules which were historical access behavior and current access environment.At the same time, the historical access behavior module was divided into two aspects: conventional behavior and gray behavior, and the current access environment module was divided into two directions: information integrity and access security.The corresponding control criteria was divided to construct the analytic network process (ANP) model under different control objectives.The limit hypermatrix under each target module was calculated to obtain the final stability weight of each element with the help of network analytic hierarchy process software.And the real user behavior data under the development platform was selected to comprehensively calculate the trust degree under different modules as the final behavior evaluation result.The expansion of the user behavior evaluation module refined the evaluation granularity, which makes the evaluation results more objective and accurate.In the cloud environment with the same malicious ratio, the optimization mechanism has better recognition effect, and it can identify cloud users with low trust efficiently and effectively, so as to improve the security and legitimacy of the cloud.At the same time, it also provids new research direction for solving the problem of user security and credibility, and effective risk control in the cloud environment.

Education and Teaching

Exploration and practice on integration of ideological, political courses into professional courses of cyberspace security specialty

Guyue LI, Aiqun HU

2022, 8(2):  183-189.  doi:10.11959/j.issn.2096-109x.2022021

Asbtract ( 415 )   HTML ( 46)   PDF (1837KB) ( 392 )   Knowledge map   

Figures and Tables | References | Related Articles | Metrics

Considering the requirements of cyberspace security talents training, this paper explored the direction and emphases of ideological and political education with the course of “new progress in cyberspace security” as an example.Ideological and political goals, elements and resources were analyzed and a “point-line-plane” scalable practical method was formed.The effectiveness of course ideological was evaluated from multiple dimensions, including the degree of integration, students’ satisfaction and sustainability.The practice results showed that teaching team and students have deeper understanding of the necessity and importance of course ideological.When students are learning these sophisticated technologies, they are also inspired by the spirit of patriotism, dedication and strive to make progress, which strengthens their determination to devote to the national cyberspace security.This course will continue integrating ideological, political courses into professional courses, optimizing the political target and teaching methods.It adapts to the demands of reformation and development in higher education and provides a useful reference for related courses in cyberspace security specialty.