Article
Automation & Control Systems
Jun Liu, Jingpan Bai, Huahua Li, Bo Sun
Summary: This article highlights the importance of real-time monitoring and correction of massive data streams in the Internet of Things (IoT) and proposes a recurrent neural network model based on LSTM+ that can reduce regression errors, detect and correct abnormal data, ensuring stability and reliability in network predictions.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Automation & Control Systems
Chunyong Yin, Sun Zhang, Jin Wang, Neal N. Xiong
Summary: This article proposes an integrated model for anomaly detection, using convolutional neural network (CNN) and recurrent autoencoder as the basis, and extracting features through two-stage sliding window data preprocessing. Empirical results show that the proposed model performs better on multiple classification metrics and achieves excellent results in anomaly detection.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
P. Manickam, M. Girija, S. Sathish, Khasim Vali Dudekula, Ashit Kumar Dutta, Yasir A. M. Eltahir, Nazik M. A. Zakari, Rafiulla Gilkaramenthi
Summary: Internet of Things (IoT) technology is widely used in smart cities, but security and privacy concerns arise as the usage increases. This paper proposes a new BBODL-ADC technique for anomaly detection and classification using deep learning, with remarkable results in experiments.
ALEXANDRIA ENGINEERING JOURNAL
(2023)
Article
Chemistry, Analytical
Asima Sarwar, Abdullah M. Alnajim, Safdar Nawaz Khan Marwat, Salman Ahmed, Saleh Alyahya, Waseem Ullah Khan
Summary: The security and privacy of the Internet of Things (IoT) are key concerns. This paper proposes an improved dynamic sticky binary particle swarm optimization (IDSBPSO) method for feature selection in IoT networks, and designs an intrusion detection system (IDS) to detect malicious data traffic. The experimental evaluation shows that the proposed method has significant advantages in terms of accuracy, number of features, and computational cost.
Article
Mathematics
E. Laxmi Lydia, A. Arokiaraj Jovith, A. Francis Saviour Devaraj, Changho Seo, Gyanendra Prasad Joshi
Summary: This paper introduces a new green energy-efficient routing technique based on DL for IoT applications, which increases network span and reduces data communication through error lossy compression. By optimizing route selection and anomaly detection, the energy efficiency and detection performance are improved.
Article
Computer Science, Artificial Intelligence
Zhen Cheng, Siwei Wang, Pei Zhang, Siqi Wang, Xinwang Liu, En Zhu
Summary: Deep autoencoder-based methods are the majority of deep anomaly detection, but they may have poor performance when distinguishing anomalies from normal data. To address this issue, an Improved AutoEncoder for unsupervised Anomaly Detection (IAEAD) is proposed, which optimizes for anomaly detection tasks and learns representations that preserve the local data structure.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2021)
Article
Computer Science, Information Systems
Yi Liu, Sahil Garg, Jiangtian Nie, Yang Zhang, Zehui Xiong, Jiawen Kang, M. Shamim Hossain
Summary: This article proposes a communication-efficient on-device federated learning (FL)-based deep anomaly detection framework for sensing time-series data in IIoT. The framework includes an FL framework, AMCNN-LSTM model, and gradient compression mechanism, which can improve generalization ability, accurately detect anomalies, and enhance communication efficiency.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Computer Science, Information Systems
Weiping Ding, Mohamed Abdel-Basset, Reda Mohamed
Summary: Our daily lives have been greatly influenced by the Internet of Things (IoT) in recent years. While IoT brings convenience and efficiency to our lives, it also exposes devices to cyberattacks due to weak security mechanisms. This paper introduces DeepAK-IoT, a deep learning model designed to detect cyberattacks against IoT devices. It utilizes three blocks – RSR, TRB, and DB – to extract features, learn temporal representations, and classify input records. Experimental results on three public datasets demonstrate DeepAK-IoT's high accuracy in detecting cyber threats in IoT systems, making it a powerful alternative model for managing cybersecurity in IoT networks.
INFORMATION SCIENCES
(2023)
Review
Chemistry, Multidisciplinary
Redhwan Al-amri, Raja Kumar Murugesan, Mustafa Man, Alaa Fareed Abdulateef, Mohammed A. Al-Sharafi, Ammar Ahmed Alkahtani
Summary: The paper discusses the importance and challenges of anomaly detection in IoT data processing, summarizing the various aspects of issues and core challenges in IoT data processing, including data characteristics, anomaly types, learning modes, window models, and more.
APPLIED SCIENCES-BASEL
(2021)
Review
Computer Science, Theory & Methods
Arnaldo Sgueglia, Andrea Di Sorbo, Corrado Aaron Visaggio, Gerardo Canfora
Summary: The widespread adoption of IoT devices has led to the automation of data collection and monitoring processes. However, this has also generated a large amount of data that needs to be managed and analyzed. To address this issue, researchers and practitioners have employed anomaly detection techniques to recognize abnormal behaviors in complex systems. In IoT environments, anomaly detection often involves the analysis of time series data under specific constraints.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2022)
Article
Automation & Control Systems
Liming Fang, Yang Li, Zhe Liu, Changchun Yin, Minghui Li, Zehong Jimmy Cao
Summary: The application of IoT in the medical field has brought unprecedented convenience but also security risks, leading to the proposal of an anomaly detection system for detecting illegal behavior (DIB) to ensure the safety of control services. The model based on rough set theory and FCVM can improve the accuracy of DIB classification anomalies.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Review
Computer Science, Information Systems
Ayan Chatterjee, Bestoun S. Ahmed
Summary: The ongoing research on anomaly detection for IoT is rapidly expanding, with a focus on areas such as network security, sensor monitoring, smart home, and smart city applications. Recent advancements have highlighted the need for further study on IoT anomaly detection applications. However, there are currently limitations in IoT anomaly detection methodologies, particularly in dealing with the integration of various sensors, data and concept drifts, and the lack of Ground Truth data.
INTERNET OF THINGS
(2022)
Article
Automation & Control Systems
Xinlei Wang, Xiaojuan Wang, Mingshu He, Min Zhang, Zikui Lu
Summary: This article proposes an attention-weighted model to enhance the detection capabilities in the widely used message queuing telemetry transport protocol in the Internet of Things. The model extracts spatial-temporal features by constructing perception node collection graphs, utilizing message-passing mechanism, bidirectional long short-term memory model, and self-attention mechanism. Experimental results demonstrate its effectiveness and high accuracy on multiple datasets.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Chemistry, Analytical
Abebe Diro, Naveen Chilamkurti, Van-Doan Nguyen, Will Heyne
Summary: The Internet of Things (IoT) is made up of a massive number of smart devices capable of data collection, storage, processing, and communication, which provides tremendous innovation opportunities but also raises concerns due to inherent vulnerabilities. The challenging IoT environment requires anomaly detection monitoring systems for better security, and blockchain-based systems can collaboratively learn to detect anomalies effectively.
Article
Chemistry, Analytical
Simona Cicero, Massimo Guarascio, Antonio Guerrieri, Simone Mungari
Summary: In recent years, the widespread use of smart devices in various types of buildings has been made possible by advancements in sensor, communication, and data storage technologies. These devices have the potential to enhance important aspects of life within buildings, including energy efficiency, safety, health, and occupant comfort. The rise of the Internet of Things has resulted in an exponential growth in connected devices and data exchange, leading to the need for new computing paradigms like Edge Computing and Fog Computing. Advanced Artificial Intelligence and Machine Learning can play a crucial role in analyzing data and predicting unexpected events in this complex scenario. However, there is currently a lack of Deep-Learning-based approaches specifically designed for ensuring safety in IoT-Based Smart Buildings. This paper proposes an unsupervised neural architecture for detecting anomalies in such contexts, which has been demonstrated to be effective in a real-world case study.
Article
Engineering, Multidisciplinary
Zia Ullah, Hasan Saeed Qazi, Ahmad Alferidi, Mohammed Alsolami, Badr Lami, Hany M. Hasanien
Summary: This study presents a novel method for optimizing energy trading within microgrids by using a hybrid of particle swarm optimization and gravitational search algorithms. The proposed approach promotes cooperative energy trading among microgrids and the main grid, considering network constraints and the uncertainty of renewable energy. Simulation results show that this method maximizes renewable energy utilization, reduces load burden on the main grid, and significantly decreases energy costs.
ALEXANDRIA ENGINEERING JOURNAL
(2024)
Article
Engineering, Multidisciplinary
Chin Joo Tan
Summary: In this study, the effect of mesh sensitivity on the hole-flanging process was investigated by varying the mesh layouts and punch surface meshing techniques. The results showed that the punch displacement and mesh parameters have significant effects on the wall thickness distributions and forming load profiles. Additionally, calibrating the simulation model's stiffness with the experimental peak load through matching the simulated peak load with the experimental peak under stability conditions was recommended.
ALEXANDRIA ENGINEERING JOURNAL
(2024)
Article
Engineering, Multidisciplinary
Wei-Chao Yang, Lun Zhao, E. Deng, Yi-Qing Ni, Wen Zhao, Yi-Kang Liu, De-Hui Ouyang
Summary: This paper establishes a three-dimensional coupled train-subgrade-wind dynamics model, and investigates the aerodynamic load variation and flow field mechanisms when high-speed trains transit different types of subgrade-cutting transition sections in crosswind conditions. The results indicate that the aerodynamic performance of the train deteriorates in these transition sections, and the aerodynamic load of the head car varies in different operating scenarios.
ALEXANDRIA ENGINEERING JOURNAL
(2024)
Article
Engineering, Multidisciplinary
Shashi Bhushan, Anoop Kumar, Eslam Hussam, Manahil SidAhmed Mustafa, Mohammed Zakarya, Wedad R. Alharbi
Summary: In the sample survey theory, accurate estimation of parameters is essential for survey practitioners. This paper suggests optimal classes of estimators by modifying conventional estimators under stratified ranked set sampling (SRSS). The suggested estimators have been shown to outperform traditional estimators, particularly regression (BLU) estimators, both theoretically and experimentally.
ALEXANDRIA ENGINEERING JOURNAL
(2024)
Article
Engineering, Multidisciplinary
Laila A. Al-Essa, Ahmed A. Soliman, Gamal A. Abd-Elmougod, Huda M. Alshanbari
Summary: In this study, the class of lifetime distributions with bathtub-shaped failure rate functions is examined. Statistical inference methods are used to estimate the parameters of the model, and the Bayesian approach is compared with classical methods. The study also discusses the evaluation of product relative merits based on lifetime duration under a hybrid censoring scheme.
ALEXANDRIA ENGINEERING JOURNAL
(2024)
Article
Engineering, Multidisciplinary
Mohammad Partohaghighi, Marzieh Mortezaee, Ali Akguel, Ahmed M. Hassan, Necibullah Sakar
Summary: The study introduces a new variant of the fractal-fractional diffusion equation using the fractal-fractional operator. It proposes a novel operational matrix technique to solve the equation, transforming it into an algebraic system. The study presents graphical and tabular representations of exact and approximated solutions, along with corresponding errors, and conducts comparative analysis of solutions at specific time points.
ALEXANDRIA ENGINEERING JOURNAL
(2024)