检索结果-内蒙古大学图书馆

CODE-SMASH: Source-Code Vulnerability detection Using Siamese and Multi-Level Neural Architecture

IEEE ACCESS 2024年 12卷 102492-102504页

作者： Han, Sungmin Nam, Hyunkyung Kang, Jaesik Kim, Kwangsoo Cho, Seungjae Lee, Sangkyun Korea Univ Sch Cybersecur Seoul 02841 South Korea LIG Nex1 Cyber Warfare Res & Dev Lab Seongnam 13488 South Korea

The rapid evolution of software development, propelled by competitive demands and the continuous integration of new features, frequently leads to inadvertent security oversights. Traditional security practices, often reactive in nature, primarily focus on identifying known vulnerabilities, creating a significant shortfall in detecting emergent, zero-day threats. This paper introduces CODE-SMASH, a novel deep learning-based source code vulnerability detector that utilizes a Siamese neural network with a hierarchical architecture integrating BiGRU and attention mechanisms. Our experiments using real-world datasets, specifically the Chromium and Debian datasets, demonstrate CODE-SMASH's superiority over existing methods. It achieves significant improvements in detection performance across all key metrics, including accuracy, precision, recall, and F1-score, with average improvements of approximately 8.3%, 11.6%, 27.75%, and 17.7%, respectively, compared to the best-performing existing methods in our experiments. Moreover, CODE-SMASH shows its superior capability in handling complex and lengthy code sequences, with performance improvements for long-length code (60 to 80 lines) in F1 scores of 4.53 percentage points on the Chromium dataset and 5.62 percentage points on the Debian dataset compared to the second-best model's performance. We believe our research makes a significant contribution to the field of automated vulnerability detection by providing a high-precision solution to the growing challenges in software security. Furthermore, based on our findings, we anticipate that future research could enhance CODE-SMASH by expanding its generalizability to various programming languages and reducing computational demands to improve efficiency.

关键词： Source coding Codes Neural networks Vectors Computer architecture Static analysis detection algorithms Software development management Code similarity hierarchical attention network Siamese neural network source code vulnerability detection

来源：评论

学校读者我要写书评

暂无评论

A Dynamic-Watermarking-Based Cyberattack detection Framework on an LFC System With Uncertain Parameters

引用

IEEE INTERNET OF THINGS JOURNAL 2024年第14期11卷 24389-24399页

作者： Jia, Shuyu Guo, Qinglai Tsinghua Univ Dept Elect Engn State Key Lab Power Syst Beijing 100084 Peoples R China

The real-time control of modern power systems faces cyber risks owing to the deep coupling of cyber systems and physical systems. Attack detection plays an important role in cybersecurity issues. In load frequency control systems, for example, detecting whether malicious data are injected guarantees the stability of the frequency. As an effective active attack detection algorithm, the dynamic watermarking algorithm faces challenge of parameter uncertainty in real-world applications. Here, we quantitatively analyze the influence of uncertain parameters for the dynamic-watermarking-based detection algorithm. By introducing the parameter identification algorithm, we propose an active attack detection framework for LFC systems with uncertain parameters, which is constructed for an online application. The proposed framework is validated in a real-world three-region power system simulation. The results show that the introduction of the parameter identification steps ensures the validity of the detection algorithm and can effectively avoid the malfunction of the detection mechanism caused by uncertain parameters.

关键词： Watermarking Heuristic algorithms Power system dynamics Control systems detection algorithms Frequency control Load modeling Attack detection cyber-physical systems dynamic watermarking load frequency control

来源：评论

学校读者我要写书评

暂无评论

Review on the Advancements in Wind Turbine Blade Inspection: Integrating Drone and Deep Learning Technologies for Enhanced Defect detection

引用

IEEE ACCESS 2024年 12卷 33236-33282页

作者： Memari, Majid Shakya, Praveen Shekaramiz, Mohammad Seibi, Abdennour C. Masoum, Mohammad A. S. Utah Valley Univ Engn Dept Machine Learning & Drone Lab Orem UT 84058 USA

The increasing demand for wind power requires more frequent inspections to identify defects in the Wind Turbine Blades (WTBs). These defects, if not detected, can compromise the structural integrity and safety of wind turbines. As WTBs are crucial and costly components, they may suffer material degradation and fatigue failure, which affects their performance and safety. Thus, the urgency for efficient and regular monitoring to maintain their structural integrity is greater than ever. This review paper explores innovative methods in fatigue testing, damage detection, and structural reliability in WTBs, focusing on the use of recent inspection methods, including those that take advantage of drones. Drones are used to identify defects such as cracks, erosion, and coating irregularities using text high-resolution imagery with the onboard cameras. Various investigators have developed novel data-driven approaches, incorporating machine learning and deep learning, to accurately identify these defects. Although deep text learning-based image processing has been successful in other public infrastructure contexts, its application to wind turbine inspection from aerial images presents unique challenges. This paper also highlights the critical role of failure inspection in enhancing the operational integrity of WTBs, showcasing state-of-the-art deep learning techniques that are pivotal for identifying and analyzing failures in WTBs from images captured by drones. The paper provides insights into the latest developments in using drone imagery for blade defect detection, contrasting this method with traditional non-destructive techniques. This approach could significantly transform the wind energy industry by offering a more efficient, automated, and precise way of ensuring the structural health of wind turbines. Unlike previous studies that predominantly focus on isolated aspects such as inspection or fatigue, this review paper not only integrates the three major aspects of WTBs

关键词： Blades Wind turbines Erosion Skin Inspection Drones Defect detection detection algorithms Anomaly detection Fault diagnosis Feature extraction Image processing Deep learning Maintenance engineering Wind power generation Wind turbine blades defect detection drones anomaly detection fault identification feature extraction image processing deep learning aerial imagery crack detection turbine maintenance fatigue reliability

来源：评论

学校读者我要写书评

暂无评论

A Novel Approach for Spam detection Using Natural Language Processing With AMALS Models

引用

IEEE ACCESS 2024年 12卷 124298-124313页

作者： Agarwal, Ruchi Dhoot, Anshita Kant, Surya Singh Bisht, Vimal Malik, Hasmat Ansari, Md. Fahim Afthanorhan, Asyraf Hossaini, Mohammad Asef JIMS Engn Management Tech Campus Dept Comp Sci Greater Noida 201303 India Moscow Inst Phys & Technol Sch Radio Engn & Comp Technol Dept Phystech Moscow 141701 Russia Graph Era Hill Univ Dept Elect & Commun Engn Bhimtal 263136 India Univ Teknol Malaysia UTM Fac Elect Engn Dept Elect Power Engn Skudai 81310 Johor Bahru Malaysia Graph Era Dept Elect Engn Dehra Dun India Univ Sultan Zainal Abidin UniSZA Artificial Intelligence Islamic Civilizat & Sustai Kuala Nerus 21300 Terengganu Malaysia Univ Sultan Zainal Abidin UniSZA Operat Res & Management Sci Kuala Nerus 21300 Terengganu Malaysia Badghis Univ Dept Phys Badghis 3351 Afghanistan

To enhance their company operations, organizations within the industry leverage the ecosystem of big data to manage vast volumes of information effectively. To achieve this objective, it is imperative to analyze textual data while prioritizing the safeguarding of data integrity and implementing robust measures for organizing and validating data through the utilization of spam filters. Various methodologies can be employed, including Word2Vec, bag-of-words, BERT, as well as term frequency & reciprocal document frequency (TF-IDF). Nevertheless, none of these solutions effectively address the problem of data scarcity, which might lead to the existence of missing information in the collected documents. To properly address this problem, it is necessary to employ a strategy that categorizes each document based on the topic matter and uses statistical approaches for approximation. This research paper presents a novel approach for spam detection using natural language processing. The proposed strategy utilizes a least-squares model to modify themes and incorporates gradient descent and altering least-squares (i.e., AMALS) models for estimating missing data. TF-IDF and uniform-distribution methods perform the estimation. The performance evaluation reveals that the suggested technique exhibits a superior performance of 98% compared to the existing industry TF-IDF model in accurately predicting spam within big data ecosystems. By this model, the environment of an organization or a company can be saved from spamming or other attacks, which can lead to extracting their data for unauthorized users to protect the details.

关键词： Unsolicited e-mail Natural language processing Machine learning Internet of Things Biological system modeling Big Data Security Artificial intelligence detection algorithms big data machine learning spam detection

来源：评论

学校读者我要写书评

暂无评论

DB-CBIL: A DistilBert-Based Transformer Hybrid Model Using CNN and BiLSTM for Software Vulnerability detection

引用

IEEE ACCESS 2024年 12卷 64446-64460页

作者： Bahaa, Ahmed Kamal, Aya El-Rahman Fahmy, Hanan Ghoneim, Amr S. Helwan Univ Fac Comp & Artificial Intelligence Dept Informat Syst Helwan 11795 Egypt Beni Suef Univ Fac Comp & Artificial Intelligence Dept Informat Syst Bani Suwayf 62521 Egypt Helwan Univ Fac Comp & Artificial Intelligence Dept Comp Sci Helwan 11795 Egypt

Software vulnerabilities are among the significant causes of security breaches. Vulnerabilities can severely compromise software security if exploited by malicious attacks and may result in catastrophic losses. Hence, Automatic vulnerability detection methods promise to mitigate attack risks and safeguard software security. This paper introduces a novel model for automatic vulnerability detection of source code vulnerabilities dubbed DB-CBIL using a hybrid deep learning model based on Distilled Bidirectional Encoder Representations from Transformers (DistilBERT). The proposed model considers contextualized word embeddings using the language model for the syntax and semantics of source code functions based on the Abstract Syntax Tree (AST) representation. The model includes two main phases. First, using a vulnerable code dataset, the pre-trained DistilBert transformer is fine-tuned for word embedding. Second, a hybrid deep learning model detects which code functions are vulnerable. The hybrid model is built on two Deep Neural Networks (DNN). The first model is the Convolutional Neural Network (CNN), which is used for extracting features. The second model is Bidirectional-LSTM (BiLSTM), which has been used to maintain the sequential order of the data as it can handle lengthy token sequences. The utilized source code dataset is derived from the Software Assurance Reference Database (SARD) benchmark dataset. Final experimental findings show that the proposed model outperforms the state-of-the-art approaches' performance by improving precision, recall, F1-score, and False Negative Rate (FNR) by 2.41%-8.95%, 4.0%-16.28%, 1.85%-12.74%, and 18% respectively. The proposed model reports the lowest FNR in the literature, a significant achievement given the cost-based nature of vulnerability detectors.

关键词： Convolutional neural networks Codes Transformers Source coding Security Feature extraction detection algorithms Deep learning Software quality Automatic vulnerability detection BERT deep learning DistilBERT transformers

来源：评论

学校读者我要写书评

暂无评论

Infrared Small Target detection Based on Hierarchical Terrace Contrast Measure

引用

IEEE ACCESS 2024年 12卷 92268-92280页

作者： Guan, Song Zhou, Dali Wang, Xiaodong Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys Changchun 130033 Peoples R China Univ Chinese Acad Sci Beijing 100049 Peoples R China Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys CAS Key Lab On Orbit Mfg & Integrat Space Opt Syst Changchun 130033 Jilin Peoples R China

The detection of infrared small targets is a topic of significant theoretical research value and practical application potential. However, the complex and dynamic detection environment introduces challenges such as intricate backgrounds, which can greatly impact the accuracy of the detection task. The existing method has limitations including the inability to detect extremely small targets, and also cannot competent to maintain exceptional detectability and good real-time performance when the small targets are under complex background. To address these issues, this paper presents an infrared small target detection method while offering superior real-time performance and exceptional detection capability. Firstly, a detection model for omnidirectional pixel differences is formulated in this paper by calculating hierarchical terrace contrast measure (HTCM) through pixel differences between the center pixel and its neighboring pixels at various levels and directions. Then, this paper exploits the radiative properties of small infrared targets by computing the minimum pixel difference between the center pixel and its neighboring pixels at various levels to simplify the model and enhance the real-time performance of the algorithm. Experimental testing on six publicly available datasets is conducted to compare our proposed method with seven existing algorithms. The results demonstrate that our approach achieves rapid detection for both large and small targets while exhibiting superior overall detection capability along with excellent real-time performance. HTCM can fastest achieve a single frame in only 0.0403 seconds and represents a 17.8% increase compared to the second place within an average single frame time.

关键词： Real-time systems Object detection Computational modeling Three-dimensional displays Solid modeling Optics Flowcharts detection algorithms Target detection infrared small target real-time performance detection capability hierarchical terrace contrast measure

来源：评论

学校读者我要写书评

暂无评论

An Asynchronous Training-Free SSVEP-BCI detection Algorithm for Non-Equal Prior Probability Scenarios

引用

IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING 2024年 32卷 4120-4130页

作者： Wang, Junsong Cui, Yuntian Zhang, Hongxin Wu, Haolin Yang, Chen Beijing Univ Post & Telecommun Sch Elect Engn Beijing 100083 Peoples R China Tsinghua Univ Sch Med Dept Biomed Engn Beijing 100084 Peoples R China

SSVEP-based brain-computer interface (BCI) systems have received a lot of attention due to their relatively high Signal to Noise Ratio (SNR) and less training requirements. Most of the existing steady-state visual evoked potential (SSVEP) detection algorithms treat the prior probability of each alternative target being selected as equal. In this study, the prior probability distribution of alternative targets was introduced into the SSVEP recognition algorithm, and an asynchronous training-free SSVEP-BCI detection algorithm for non-equal prior probability scenarios was proposed. This algorithm is based on the Spatio-temporal equalization multi-window technique (STE-MW) and introduces the Maximum A Posteriori criterion (MAP), which makes full use of prior information to improve the performance of the asynchronous training-free BCI system. In addition, we proposed a mutual information-based performance evaluation metric called Mutual information rate (MIR) specifically for non-equal prior probability scenarios. This evaluation framework is designed to provide a more accurate estimation of the information transmission performance of BCI systems in such scenarios. A 10-target simulated vehicle control offline experiment involving 17 subjects showed that the proposed method improved the average MIR by 6.48%. Online free control experiments involving 12 subjects showed that the proposed method improved the average MIR by 14.93%, and significantly reduced the average instruction time. The proposed algorithm is more suitable for practical engineering application scenarios that are asynchronous and training-free;the extremely high accuracy is guaranteed while maintaining a low false alarm rate, which can be applied to asynchronous BCI systems that require high stability.

关键词： Electroencephalography Probability distribution Visualization detection algorithms Brain modeling Background noise Training Target recognition Posterior probability Mathematical models Brain-computer interface (BCI) steady-state visual evoked potential (SSVEP) maximum a posteriori (MAP) asynchronous mutual information rate (MIR)

来源：评论

学校读者我要写书评

暂无评论

Relaxed Co-Design of Attack detection and Set-Membership Estimation for T-S Fuzzy Systems Subject to Malicious Attacks

引用

IEEE TRANSACTIONS ON FUZZY SYSTEMS 2024年第5期32卷 2663-2676页

作者： Du, Mengni Xie, Xiangpeng Wang, Hui Xia, Jianwei Chadli, Mohammed Liaocheng Univ Sch Math Sci Liaocheng 252059 Peoples R China Nanjing Univ Posts & Telecommun Sch Internet Things Nanjing 210023 Peoples R China Univ Paris Saclay IBISC Lab F-91020 Evry France

This article is concerned with the attack detection problem of Takagi-Sugeno (T-S fuzzy) systems with unknown but bounded (UBB) noise subject to malicious attacks via the set-membership estimation and a switching multimode high-order free-weighting matrix (SMHFM) method. Initially, the SMHFM is developed to address the so-called conservatism problem caused by the one-order free-weighting matrix method. To solve this problem, the homogeneous polynomial technique is employed to introduce groups of free-weighting matrices for different switching modes. As a result, the SMHFM is synchronized with the working modes and possesses a high-order feature. In addition, a switching multimode mechanism is implemented to enable the SMHFM to exhibit multiple modes. Time-variant balanced matrices are introduced for different switching modes to adjust the relevant matrix terms. This adjustment allows for obtaining more relaxed conditions, leading to smaller constraint sets and higher accuracy in state estimation. Furthermore, a zonotope-based set-membership attack detection algorithm is introduced for T-S fuzzy systems, which is capable of detecting various types of attacks. By utilizing the proposed SMHFM method for attack detection, the level of conservatism in state estimation can be reduced. Finally, two simulation examples are given and some comparisons are made to validate the effectiveness of the proposed methods.

关键词： Fuzzy systems Switches Security State estimation detection algorithms Noise measurement Heuristic algorithms Homogeneous polynomial malicious attacks detection multimode mechanism set-membership theory

来源：评论

学校读者我要写书评

暂无评论

Community detection for Heterogeneous Multiple Social Networks

引用

IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS 2024年第5期11卷 6966-6981页

作者： Zhu, Ziqing Yuan, Guan Zhou, Tao Cao, Jiuxin China Univ Min & Technol Sch Comp Sci & Technol Xuzhou 221116 Jiangsu Peoples R China China Univ Min & Technol Jiangsu Key Lab Mine Mech & Elect Equipment Xuzhou 221116 Jiangsu Peoples R China Minist Educ Engn Res Ctr Digitizat Mine Xuzhou 221116 Jiangsu Peoples R China Nanjing Tech Univ Coll Comp & Informat Engn Nanjing 211816 Peoples R China Southeast Univ Sch Cyber Sci & Engn Nanjing 211189 Peoples R China Purple Mt Labs Nanjing 211111 Peoples R China

The community plays a crucial role in understanding user behavior and network characteristics in social networks. Some users can use multiple social networks at once for a variety of objectives. These users are called overlapping users who bridge different social networks. Detecting communities across multiple social networks is vital for interaction mining, information diffusion, and behavior migration analysis among networks. This article presents a community detection method based on nonnegative matrix trifactorization for multiple heterogeneous social networks, which formulates a common consensus matrix to represent the global fused community. Specifically, the proposed method involves creating adjacency matrices based on network structure and content similarity, followed by alignment matrices that distinguish overlapping users in different social networks. With the generated alignment matrices, the method could enhance the fusion degree of the global community by detecting overlapping user communities across networks. The effectiveness of the proposed method is evaluated with new metrics on Twitter, Instagram, and Tumblr datasets. The results of the experiments demonstrate its superior performance in terms of community quality and community fusion.

关键词： Social networking (online) Multiplexing Topology Blogs Nonhomogeneous media detection algorithms Symmetric matrices Clustering community detection data mining matrix factorization social network

来源：评论

学校读者我要写书评

暂无评论

Effect of Cyberattack on Event detection Algorithm in Distribution System Using Synchrophasor Measurements

引用

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS 2024年第5期60卷 7613-7622页

作者： Beura, Subal Padhy, Bibhu Prasad IIT Ropar Dept Elect Engn Rupnagar 140001 India

Cyber-physical systems (CPSs) in the twenty-first century are paving new ways to have better connectivity among the power system devices, effective control, and improved reliability. Thus, manipulating this system by foreign interventions like false data injection (FDI) must be handled with utmost priority. Power system monitoring using micro phasor measurement units (mu PMUs) gives more insight into the transient performance of the system. Using the optimal placement of PMUs, fewer mu PMUs can ensure the complete observability of the power system. This comprehensive view of the system can be used to detect any event that occurs in the system. The measurements of mu PMUs connected to the end points of each radial line can be used to detect the event location. However, any FDI-type attack on these devices can degrade the result of the event detection algorithm. In this paper, we work on the IEEE 33 bus system to find the effect of falsified mu PMU data on the event detection algorithm. This work presents attack detection using a tolerance limits violation in the presence of an attack. Further, attack effect mitigation on the mu PMU readings is inspected to find its detrimental effect on the detection algorithm. The manipulated mu PMUs are detected, and their readings are avoided for the proper event location. Event location information is processed to find the control action needed to maintain the stability of the power system. Further, simultaneous event detection of multiple events using a transformer neural network (TNN) classifier is implemented to improve the resiliency of smartgrids.

关键词： Phasor measurement units Event detection Voltage Power system stability Switches detection algorithms Cyberattack Cyber-physical system false data injection attack event detection neural network real time system

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：