Article
Engineering, Electrical & Electronic
Sravanthi Godala, M. Sunil Kumar
Summary: This paper proposes a new hierarchical intrusion detection system for accurately determining malicious sensor nodes. The system operates in two levels and incorporates optimization techniques to improve prediction accuracy, performing better compared to other modernization models.
OPTICAL AND QUANTUM ELECTRONICS
(2023)
Review
Computer Science, Artificial Intelligence
Insoo Sohn
Summary: With the increase in IoT devices, the amount of personal and sensitive data flowing through global networks has grown rapidly, making cybersecurity a crucial issue for future network evolution. Deep learning techniques, particularly deep belief networks (DBN), have become key solutions in detecting malicious attacks.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Theory & Methods
Ying Liu, Ting Zhi, Ming Shen, Lu Wang, Yikun Li, Ming Wan
Summary: This paper proposes a two-level DDoS attack detection method based on information entropy and deep learning to effectively detect attack traffic in the SDN environment.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2022)
Article
Computer Science, Information Systems
Murtaza Ahmed Siddiqi, Wooguil Pak
Summary: This paper presents a method that combines image processing with convolutional neural networks (CNN) for network intrusion detection systems (NIDS). The method converts non-image data from network traffic into images, enhances them using Gabor filters, and classifies them using a CNN classifier. Through comparisons with benchmark datasets, the proposed method demonstrates higher precision.
CMC-COMPUTERS MATERIALS & CONTINUA
(2022)
Article
Computer Science, Artificial Intelligence
Wei Song, Shiyu Zhang, Zijian Wen, Junhao Zhou
Summary: This study proposed a novel adaptive learning deep belief network (ALDBN) that dynamically adjusts its structure for feature extraction using a series of growing and pruning algorithms. It also revealed the relationships between network depth, information entropy, and weight distribution, while providing theoretical proof for convergence and comparing performance with other methods on benchmark datasets. The results show that ALDBN outperforms competitors in terms of accuracy on various tests.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Geeta Kocher, Gulshan Kumar
Summary: The paper provides a comprehensive overview of the applications and research status of machine learning methods and deep learning methods in intrusion detection. It discusses their performance, advantages, and experimental results. Moreover, it also explores the current research challenges and issues in the field, aiming to assist fellow researchers in the area.
Article
Engineering, Civil
Guoqi Xie, Laurence T. Yang, Yuanda Yang, Haibo Luo, Renfa Li, Mamoun Alazab
Summary: With the advancement of Internet of vehicles and autonomous driving technologies, automotive Controller Area Networks (CAN) face various security threats. A enhanced deep learning GAN model is proposed to address this issue, which improves detection accuracy of data tampering threat by designing elaborate CAN message blocks and enhancing GAN discriminator.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Engineering, Mechanical
Zhiwu Shang, Wanxiang Li, Maosheng Gao, Xia Liu, Yan Yu
Summary: This paper proposes a multi-scale deep feature fusion intelligent fault diagnosis method based on information entropy, using a variety of autoencoders to construct a deep neural network feature extraction structure and employing deep belief network probability model as the fault classifier. Experimental results show that compared to traditional methods, this approach obtains higher accuracy features from raw data.
CHINESE JOURNAL OF MECHANICAL ENGINEERING
(2021)
Article
Computer Science, Interdisciplinary Applications
Mariya Princy Antony Saviour, Dhandapani Samiappan
Summary: Network security has been improved by intrusion detection, which can identify unexpected threats in network traffic. Modern methods for detecting network anomalies rely on traditional machine learning models. However, these systems depend on human-designed traffic features that are no longer relevant in the era of big data, leading to lower accuracy and certain exceptional features. This research aims to develop a storage authentication and access control model based on the Interplanetary File System (IPFS) and a network intrusion detection system based on Chronological Anticorona Virus Optimization (CACVO-based DRN).
ADVANCES IN ENGINEERING SOFTWARE
(2023)
Article
Computer Science, Information Systems
S. Rajasoundaran, S. V. N. Santhosh Kumar, M. Selvi, K. Thangaramya, Kannan Arputharaj
Summary: Underwater Wireless Sensor Networks (UWSNs) transmit data through water medium and monitor oceanic conditions, under-sea habitations, and military objects. Protecting the vulnerable UWSN channel under critical water conditions is challenging. This article proposes a new intrusion detection system with integrated secure MAC principles and LSTM architectures for real-time neighbor monitoring tasks.
Article
Engineering, Marine
Tianhao Hou, Hongyan Xing, Xinyi Liang, Xin Su, Zenghui Wang
Summary: Marine sensors are vulnerable to illegal access network attacks, and the nation's meteorological and hydrological information faces increasing risk. This paper proposes a deep learning-based network intrusion detection method, combining LCVAE with CBiLSTM to generate virtual samples and extract significant attack features from traffic data.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Mathematics, Applied
Hao Liu, Langzhou He, Fan Zhang, Zhen Wang, Chao Gao
Summary: This paper proposes an optimized dynamic deep graph infomax (ODDGI) method for dynamic community detection, which utilizes recurrent neural network (RNN) to capture the dynamism of networks while avoiding storing all information of dynamic networks. The method considers the importance of nodes using similarity aggregation strategy to improve the accuracy of node representation.
Article
Computer Science, Information Systems
Bishwajeet Kumar Pandey, M. R. M. Veeramanickam, Shabeer Ahmad, Ciro Rodriguez, Doris Esenarro
Summary: The evolution of the Internet has led to an abundance of information, making the internet world more complex and susceptible to powerful attacks. The Intrusion Detection System (IDS) plays a crucial role in network security in modern networks, with two types of IDS - anomaly-based and signature-based behavior detection. Researchers have proposed various detection approaches for network intrusions. This paper introduces a deep learning approach using the Exponential Shuffled Shepherded Optimization Algorithm (ExpSSOA) for intrusion detection. The proposed ExpSSOA-based Deep Maxout network is evaluated using the MQTT-IOT-IDS2020 dataset and the Apache Web Server dataset, showing better results in terms of accuracy, F-measure, precision, and recall.
COMPUTERS & SECURITY
(2023)
Article
Computer Science, Hardware & Architecture
Jieling Li, Hao Zhang, Yanhua Liu, Zhihuang Liu
Summary: This paper proposes a semi-supervised machine learning framework for network intrusion detection, which combines multi-strategy feature filtering, PCA, and an improved Tri-LightGBM model based on stratified sampling to enhance detection accuracy and classification performance.
JOURNAL OF SUPERCOMPUTING
(2022)
Article
Engineering, Civil
Zhongru Wang, Xinzhou Xie, Lei Chen, Shouyou Song, Zhongjie Wang
Summary: The study aims to enhance the safety performance of urban construction in the traffic field by utilizing the deep convolution neural network model AlexNet with more network layers and stronger learning ability, and introducing the Gate Recurrent Unit (GRU) neural network into the improved AlexNet to construct an intrusion detection model for the urban rail transit management system. By verifying the model performance with collected data and simulation experiments, it achieves a recognition accuracy of 96.00% for intrusion detection, which is at least 1.55% higher than other neural network models. Furthermore, the model demonstrates stable training and test times as well as high data transmission security performance, providing an experimental basis for improving the safety performance of rail transit systems in smart cities.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Review
Computer Science, Information Systems
Euclides Carlos Pinto Neto, Sajjad Dadkhah, Somayeh Sadeghi, Heather Molyneaux, Ali A. Ghorbani
Summary: The Internet of Things (IoT) has the potential to revolutionize medical treatment in healthcare, but it also faces security threats. Advanced analytics can enhance IoT security, but generating realistic datasets is complex. This research conducts a review of Machine Learning (ML) solutions for IoT security in healthcare, focusing on existing datasets, resources, applications, and challenges, to highlight the current landscape and future requirements.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Duncan Deveaux, Takamasa Higuchi, Seyhan Ucar, Jerome Harri, Onur Altintas
Summary: This paper investigates the ability to predict the risk patterns of vehicles in a roundabout and suggests that constraining knowledge transfer to roundabouts with a similar context can significantly improve accuracy.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Lingjun Zhao, Qinglin Yang, Huakun Huang, Longtao Guo, Shan Jiang
Summary: Metaverse seamlessly integrates the real and virtual worlds, and intelligent wireless sensing technology can serve as an intelligent, flexible, non-contact way to access the metaverse and accelerate the establishment of a bridge between the real physical world and the metaverse. However, there are still challenges and open issues in this field.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Jing Xiong, Hong Zhu
Summary: With the rapid growth of data in the era of IoT, the challenge of data privacy protection arises. This article proposes a federated learning approach that uses collaborative training to obtain a global model without direct exposure to local datasets. By utilizing dynamic masking and adaptive differential privacy methods, the approach reduces communication overhead and improves the converge performance of the model.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Carlos Rubio Garcia, Simon Rommel, Sofiane Takarabt, Juan Jose Vegas Olmos, Sylvain Guilley, Philippe Nguyen, Idelfonso Tafur Monroy
Summary: The reliance on asymmetric public key cryptography and symmetric encryption for cyber-security in current telecommunication networks is threatened by quantum computing technology. Quantum Key Distribution and post-quantum cryptography provide resistance to quantum attacks. This paper proposes two novel hybrid solutions integrating QKD and PQC into TLS for quantum-resistant key exchange.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Annisa Sarah, Gianfranco Nencioni
Summary: This article explores the concept of a Slice Broker, an intermediate entity that purchases resources from Infrastructure Providers to offer customized network slices to users. The article proposes a cost-minimization problem and compares it with alternative problems to demonstrate its effectiveness and cost-saving capabilities.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Sumana Maiti, Sudip Misra, Ayan Mondal
Summary: The broadcast proxy re-encryption methods extend traditional proxy re-encryption mechanisms and propose a scheme called MBP for IoT applications. MBP calculates a single re-encryption key for all user groups and uses multi-channel broadcast encryption to reduce security element size. However, it increases computation time for receiver IoT devices. The use of Rubinstein-Stahl bargaining game approach addresses this issue and MBP is secure against selective group chosen-ciphertext attack in the random oracle model.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Pankaj Kumar, Hari Om
Summary: This paper presents NextGenV2V, a protocol for the next-generation vehicular network that achieves authenticated communication between vehicles using symmetric keys and a (2, n)-threshold scheme. The protocol reduces communication overhead and improves authentication delay, ensuring better security. Comparative analysis demonstrates the suitability of NextGenV2V in next-generation vehicular networks.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Eric Ossongo, Moez Esseghir, Leila Merghem-Boulahia
Summary: The implementation of 5G networks allows for the efficient coexistence of heterogeneous services in a single physical virtualized infrastructure. Virtualization of network functions enables more flexible resource management and customizable services. However, the increasing number of connected objects poses challenges in managing physical and virtual resources, requiring intelligent systems to ensure communication quality.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Suvrima Datta, U. Venkanna
Summary: The Internet of Things (IoT) enables real-time sensing and data transmission to make homes smarter. Effective device-type identification methods are crucial as the number of IoT devices continues to grow. In this paper, a P4-based gateway called PiGateway is proposed to classify and prioritize the type of IoT devices. By utilizing a decision tree model and flow rules, PiGateway enables real-time granular analysis and in-network classification of IoT traffic.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Fahad Razaque Mughal, Jingsha He, Nafei Zhu, Saqib Hussain, Zulfiqar Ali Zardari, Ghulam Ali Mallah, Md. Jalil Piran, Fayaz Ali Dharejo
Summary: This paper explores the relationship between heterogeneous cluster networks and federated learning, as well as the challenges of implementing federated learning in heterogeneous networks and the Internet of Things. The authors propose an Intra-Clustered FL (ICFL) model that optimizes computation and communication to select heterogeneous FL nodes in each cluster, enabling efficient processing of asynchronous data and ensuring data security.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Rajesh Kumar, Deepak Sinwar, Vijander Singh
Summary: This paper investigates the coexistence mechanisms between eMBB and URLLC traffic for resource scheduling in 5G. Through examining different approaches and performance metrics, it provides detailed insights for researchers in the field, and highlights key issues, challenges, and future directions.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Giovanni Nardini, Giovanni Stea
Summary: Digital Twins of Networks (DTNs) are proposed as digital replicas of physical entities, enabling efficient data-driven network management and performance-driven network optimization. DTNs provide simulation services for dynamic reconfiguration and fault anticipation, using discrete-event network simulators as the ideal tools. Challenges include centralized vs. distributed implementation, input gathering from the physical network, security issues and hosting. The possibilities of network simulation for what-if analysis are explored, with the concepts of lockstep and branching analysis defined.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Zhaolin Ma, Jiali You, Haojiang Deng
Summary: This paper presents the Distributed In-Network Name Resolution System (DINNRS), which leverages software-defined networking and Information-Centric Networking (ICN) paradigm to provide high scalability and minimal request delay. Our methods, including an enhanced marked cuckoo filter for fast resolving, achieve significant performance gains in simulation experiments.
COMPUTER COMMUNICATIONS
(2024)
Article
Computer Science, Information Systems
Yujie Wang, Ying Wang, Qingqing Liu, Yong Zhang
Summary: This paper proposes a dynamic indoor positioning method based on multi-scale metric learning of the channel state information (CSI). By constructing few-shot learning tasks, this method can achieve dynamic positioning using CSI signals without additional equipment. Experimental results show that compared to commonly used dynamic location and tracking algorithms, the proposed method has higher positioning accuracy and does not accumulate errors.
COMPUTER COMMUNICATIONS
(2024)