检索结果-内蒙古大学图书馆

Detection and Prevention of Black Hole Attack using Tree Hierarchical Deep Convolutional Neural Network and Enhanced Identity-based Encryption in Vehicular Ad Hoc Network

引用

IEIE Transactions on Smart Processing and Computing 2024年第1期13卷 41-50页

作者： Narayanan, K. Lakshmi Naresh, R. Department of Computer Science and Engineering SRM Institute of Science and Technology Kattankulathur Chengalpattu Tamilnadu Chennai603203 India Department of Networking and Communications SRM Institute of Science and Technology Kattankulathur Chengalpattu Tamilnadu Chennai603203 India

A method for the recognition and prevention of a black hole attack is proposed using a tree hierarchical deep convolutional neural network (THDCNN)and enhanced identity based encryption in a vehicular ad hoc network (VANET). Automobiles are organized with a cluster formula using an improved k-means clustering algorithm. Following the cluster formation, cluster head (CH) selection is done using a balancing composite motion optimization (BCMO) algorithm. After selecting the cluster head, the entrance of a spiteful node occurs in the cluster. The THDCNN is proposed for classifying a cluster node as i) a black hole attack node or ii) a normal node. If a black hole attack node is found, the attack node information is communicated to an individual CH, which makes a final decision. Otherwise, the standard node data is encrypted with the enhanced identity-based encryption algorithm. Finally, the proposed method prevents the normal node data from the attacker. Thus, the proposed method attains higher accuracy and lower computational time than other methods. Copyrights © 2024 The institute of Electronics and Information Engineers.

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

Classification of Conversational Sentences Using an Ensemble Pre-Trained Language Model with the Fine-Tuned Parameter

引用

computers, Materials & Continua 2024年第2期78卷 1669-1686页

作者： R.Sujatha K.Nimala Department of Networking and Communications SRM Institute of Science and TechnologyKattankulathurChengalpattuTamilnadu603203India

Sentence classification is the process of categorizing a sentence based on the context of the *** categorization requires more semantic highlights than other tasks,such as dependence parsing,which requires more syntactic *** existing strategies focus on the general semantics of a conversation without involving the context of the sentence,recognizing the progress and comparing *** ensemble pre-trained language model was taken up here to classify the conversation sentences from the conversation *** conversational sentences are classified into four categories:information,question,directive,and *** classification label sequences are for analyzing the conversation progress and predicting the pecking order of the *** of Bidirectional Encoder for Representation of Transformer(BERT),Robustly Optimized BERT pretraining Approach(RoBERTa),Generative Pre-Trained Transformer(GPT),DistilBERT and Generalized Autoregressive Pretraining for Language Understanding(XLNet)models are trained on conversation corpus with *** tuning approach is carried out for better performance on sentence *** Ensemble of Pre-trained Language Models with a Hyperparameter Tuning(EPLM-HT)system is trained on an annotated conversation *** proposed approach outperformed compared to the base BERT,GPT,DistilBERT and XLNet transformer *** proposed ensemble model with the fine-tuned parameters achieved an F1_score of 0.88.

关键词： Bidirectional encoder for representation of transformer conversation ensemble model fine-tuning generalized autoregressive pretraining for language understanding generative pre-trained transformer hyperparameter tuning natural language processing robustly optimized BERT pretraining approach sentence classification transformer models

来源：评论

学校读者我要写书评

暂无评论

Domain generalization with semi-supervised learning for people-centric activity recognition

引用

science China(Information sciences) 2025年第1期68卷 171-188页

作者： Jing LIU Wei ZHU Di LI Xing HU Liang SONG Academy for Engineering & Technology Fudan University Shanghai East-bund Research Institute on Networking Systems of AI School of Optoelectronic Information and Computer Engineering University of Shanghai for Science & Technology

People-centric activity recognition is one of the most critical technologies in a wide range of real-world applications,including intelligent transportation systems, healthcare services, and brain-computer interfaces. Large-scale data collection and annotation make the application of machine learning algorithms prohibitively expensive when adapting to new tasks. One way of circumventing this limitation is to train the model in a semi-supervised learning manner that utilizes a percentage of unlabeled data to reduce the labeling burden in prediction tasks. Despite their appeal, these models often assume that labeled and unlabeled data come from similar distributions, which leads to the domain shift problem caused by the presence of distribution gaps. To address these limitations, we propose herein a novel method for people-centric activity recognition,called domain generalization with semi-supervised learning(DGSSL), that effectively enhances the representation learning and domain alignment capabilities of a model. We first design a new autoregressive discriminator for adversarial training between unlabeled and labeled source domains, extracting domain-specific features to reduce the distribution gaps. Second, we introduce two reconstruction tasks to capture the task-specific features to avoid losing information related to representation learning while maintaining task-specific consistency. Finally, benefiting from the collaborative optimization of these two tasks, the model can accurately predict both the domain and category labels of the source domains for the classification task. We conduct extensive experiments on three real-world sensing datasets. The experimental results show that DGSSL surpasses the three state-of-the-art methods with better performance and generalization.

关键词： activity recognition deep learning domain generalization semi-supervised learning adversarial training

来源：评论

学校读者我要写书评

暂无评论

Fake news detection on social media using Adaptive Optimization based Deep Learning Approach

引用

Multimedia Tools and Applications 2025年第9期84卷 6091-6111页

作者： Steni Mol, T.S. Sreeja, P.S. Department of Computer Science Hindustan Institute of Technology and Science Chennai India Department of Computer Applications Vellore Institute of Technology Chennai India

Various content-sharing platforms and social media are developed in recent times so that it is highly possible to spread fake news and misinformation. This kind of news may cause chaos and panic among people. The automatic and accurate detection of fake news is a complex task. To consider this issue, an automatic and effective fake news detection approach is implemented. The proposed methodology is named Adaptive Adam Adadelta Optimizer-based Deep Bi-directional Long Short Term Memory (Bi-LSTM), in which Adaptive Adam AO is the merging of Adadelta Optimizer (AO), Adaptive concept, and Adam Optimization. The Bidirectional Encoder Representations from Transformers (BERT) is used for the tokenization and then, feature mining is done. Also, the dimensionality of the features is reduced by Singular Value Decomposition (SVD) and Moving Principal Component Analysis (MPCA). From them, the top-N feature selection is performed using Joint Quantum Entropy. Finally, for fake news detection, Adaptive Adam AO trained Deep Bi-LSTM is used, which obtained maximum specificity, sensitivity, and accuracy of 0.928, 0.945, and 0.938. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Hybrid Seagull and Whale Optimization Algorithm-Based Dynamic Clustering Protocol for Improving Network Longevity in Wireless Sensor Networks

引用

China Communications 2024年第10期21卷 113-131页

作者： P.Vinoth Kumar K.Venkatesh Department of Networking and Communications School of ComputingCollege of Engineering and TechnologySRM Institute of Science and TechnologyKattankulathur 603203India

Energy efficiency is the prime concern in Wireless Sensor Networks(WSNs) as maximized energy consumption without essentially limits the energy stability and network lifetime. Clustering is the significant approach essential for minimizing unnecessary transmission energy consumption with sustained network lifetime. This clustering process is identified as the Non-deterministic Polynomial(NP)-hard optimization problems which has the maximized probability of being solved through metaheuristic *** adoption of hybrid metaheuristic algorithm concentrates on the identification of the optimal or nearoptimal solutions which aids in better energy stability during Cluster Head(CH) selection. In this paper,Hybrid Seagull and Whale Optimization Algorithmbased Dynamic Clustering Protocol(HSWOA-DCP)is proposed with the exploitation benefits of WOA and exploration merits of SEOA to optimal CH selection for maintaining energy stability with prolonged network lifetime. This HSWOA-DCP adopted the modified version of SEagull Optimization Algorithm(SEOA) to handle the problem of premature convergence and computational accuracy which is maximally possible during CH selection. The inclusion of SEOA into WOA improved the global searching capability during the selection of CH and prevents worst fitness nodes from being selected as CH, since the spiral attacking behavior of SEOA is similar to the bubble-net characteristics of WOA. This CH selection integrates the spiral attacking principles of SEOA and contraction surrounding mechanism of WOA for improving computation accuracy to prevent frequent election process. It also included the strategy of levy flight strategy into SEOA for potentially avoiding premature convergence to attain better trade-off between the rate of exploration and exploitation in a more effective manner. The simulation results of the proposed HSWOADCP confirmed better network survivability rate, network residual energy and network overall throughput on par wi

关键词： clustering energy stability network lifetime seagull optimization algorithm(SEOA) whale optimization algorithm(WOA) wireless sensor networks(WSNs)

来源：评论

学校读者我要写书评

暂无评论

Mode Management of Peripherals Based on State Transition Model in FRP Language for Embedded Systems

Computer Software

引用

computer Software 2025年第1期42卷 40-53页

作者： Takimoto, Satoshi Moriguchi, Sosuke Watanabe, Takuo Department of Computer Science Tokyo Institute of Technology Institute of Science Tokyo Japan

XStorm, an FRP language for small-scale embedded systems, allows us to concisely describe state-dependent behaviors based on the state transition model. However, when we use different sets of peripheral devices depending on states, device management, such as switching power modes, should be implemented in a driver code in C. This would result in bugs as inconsistency between the state in the XStorm program and that in the driver code cannot be detected. In this research, we extend XStorm’s state hook model to express modes of peripherals that depend on states. By the extension, the language manages modes of peripherals, and thus the inconsistency is statically avoided. © 2025 Japan Society for Software science and technology. All rights reserved.

关键词： C (programming language)

来源：评论

学校读者我要写书评

暂无评论

A new perspective exploration of machine learning algorithms for defending Side-Channel attacks

引用

Multimedia Tools and Applications 2025年第16期84卷 17021-17040页

作者： Ramakrishnan, Vedhavathy Thoguluva Seethapathy, Murugaanandam Sundar, Ramesh Dhandapani, Saveetha Munusamy, Sundarrajan Balasubrahmaniam, Lakshmi Dhevi Department of Networking and Communications School of Computing Faculty of Engineering and Technology SRM Institute of Science and Technology Kattankulathur India

Machine learning algorithms are used in various real-time applications, where security is one of the major problems. Security is applied in various aspects of the application in cloud computing. One of the security issues is a side-channel attack, which is not a common problem that can be solved by a point solution. The emerging computing industries need a better solution for side-channel attack detection since the amount of data used in communication and sharing on the Internet is enormous. Several old research works have concluded that machine learning algorithms are highly suitable for data analytics. Hence, this paper has aimed to propose a better machine-learning model for side-channel attack detection. This paper explains some machine learning algorithms from a technical point of view. The explanation identifies that the Convolution Neural Network algorithm is selected as a better algorithm for side-channel attack detection. The proposed CNN model has experimented with Python software to verify the results. Since CNN is more efficient in data analytics, the analysis result produces more accurate predictions than earlier algorithms. The performance of the proposed CNN is compared with the earlier algorithms for evaluating it. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Data Analytics

来源：评论

学校读者我要写书评

暂无评论

Block chain-based trusted smart contract for secure mobile user payment gateway in e-health systems

引用

Soft Computing 2024年第17-18期28卷 10139-10151页

作者： Saranya, A. Naresh, R. Department of Computer Science and Engineering SRM Institute of Science and Technology Kattankulathur Chengalpattu Tamil Nadu Chennai603203 India Department of Networking and Communications SRM Institute of Science and Technology Kattankulathur Chengalpattu Tamil Nadu Chennai603203 India

In other instances, the cloud transfers payment information directly to the merchant’s server without first doing any fraud checks. Block authentication between the cloud and a healthcare merchant server is the goal of this study, which is designed to prevent the deployment of fraudulent servers. Smart contract verification on the blockchain will be used to this end. The cloud assault must be prevented from accessing personal payment and card information. It is via the employment of cryptography methods that this concept of privacy reservation is made possible. The Boolean crossover depth first search (DFS) technique is used to encrypt data in the payment gateway protocol at its inception. When the hospital administrator solicits card data from patients or users, our innovative approach swiftly initiates data encryption using the DFS technique. This method meticulously traverses the data structure in a depth-first manner, employing Boolean operations to systematically encrypt sensitive information. Through this process, the patient's payment and card data undergo secure transformation, rendering it unreadable to unauthorized entities. Additionally, an authentication key is generated concurrently with the encryption process, enabling verification between the system and the user before any data transmission occurs. The integration of the DFS technique serves as a fundamental layer of defense against potential cloud-based attacks. Its application ensures the utmost protection of personal payment and card details throughout the transaction process, bolstering the security infrastructure of our proposed system. A comprehensive analysis comparing our approach with existing methods showcases the efficiency and reliability of our proposed system in terms of Execution Time (ms), Encryption Time (ms), Decryption Time (ms), and Memory (bits). This study aims to bridge the gap in transactional security, ensuring robust protection against unauthorized access and data breaches in

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

Using DTL-MD with GANs and ResNet for Malicious Code Detection

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第14期49卷 63-78页

作者： Li, Yiming Xie, Tao Mei, Dongdong Department of Computer Science and Technology School of Computer Science and Engineering Ningxia Institute of Science and Technology Shizuishan753000 China

This study proposes a malicious code detection model DTL-MD based on deep transfer learning, which aims to improve the detection accuracy of existing methods in complex malicious code and data scarcity. In the feature extraction process, the weighted sum method of GIST and LBP features is used to combine the advantages of the two features. Online transfer learning is used to reduce the data distribution difference between the target domain and the source domain. The model uses ResNet50V2 as the backbone network and combines SimAM to enhance the feature extraction and representation capabilities. In addition, in order to further improve the robustness of detection, GAN is used to generate malicious code variants and expand the training data set. In the experiment, the public CICIDS 2017 data set is used for model training and testing. The performance test results show that when the threshold is 0.7, the accuracy of DTL-MD is 95.8% and the F1 score is 0.93. In a performance test involving 30,000 samples, the throughput of the DTL-MD model under Trojans, viruses, worms, and adware is 11, 12, 11, and 12 tasks/s, respectively, and the inference time is 211, 225, 239, and 234 samples/s, respectively. Compared with GAN, DTL-MD increases the throughput by about 10% and the inference speed by about 15%. The research aims to provide new ideas for improving the intelligence and automation level of malicious code detection technology, which has certain application value and practical significance. © 2025 Slovene Society Informatika. All rights reserved.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Flameshift Protocol: Revolutionizing Interoperability with Dynamic Asset Recycling for Cross-Chain Communications

引用

SN computer science 2024年第6期5卷 773页

作者： Vijayalakshmi, C. Florence, S. Metilda Department of Networking and Communications SRM Institute of Science and Technology

The burgeoning landscape of blockchain technology has ushered in an era of unprecedented innovation and opportunity. However, one of the most persistent challenges that emerge in this decentralized ecosystem is the lack of efficient and secure interoperability mechanisms across different blockchain networks. The FlameShift Protocol introduces a ground-breaking approach to address this challenge, enabling seamless cross-chain communications through a novel Dynamic Asset Recycling (DAR) mechanism. This paper delves into the intricacies of the FlameShift Protocol, illustrating its capability to not only facilitate secure and efficient cross-chain transactions but also to enhance the liquidity and utility of digital assets across disparate blockchain ecosystems. At the heart of this protocol is the DAR mechanism, which employs a unique auto-burn and reclaim process. This process ensures that when assets are transferred from one chain to another, they are temporarily locked in the source chain, and an equivalent value is minted in the destination chain. This mechanism is designed to maintain a balanced ledger across chains without the need for centralized intermediaries, thereby mitigating the risk of double-spending and other fraudulent activities. Furthermore, the FlameShift Protocol incorporates advanced cryptographic techniques and smart contract functionalities to ensure that the asset transfer process is not only secure but also transparent and verifiable. This paper presents a comprehensive analysis of the protocol’s architecture, security features, and the potential implications for the future of blockchain interoperability. By offering a scalable, secure, and efficient framework for cross-chain communications, the FlameShift Protocol paves the way for a more interconnected and versatile blockchain ecosystem. This study aims to contribute to the ongoing discourse on blockchain interoperability, providing insights and practical solutions for developers, researcher

关键词： Blockchain technology Cross-chain communications Dynamic asset recycling (DAR) FlameShift protocol Interoperability mechanisms

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：