Article
Computer Science, Artificial Intelligence
Amir Basati, Mohammad Mehdi Faghih
Summary: The Internet of Things (IoT) is widely used in various fields such as the control industry, industrial plants, and medicine, and security implementation is crucial. Network intrusion detection systems (NIDSs) play an important role in detecting network attacks and threats. However, current NIDSs based on deep learning (DL) models require a large number of processing resources due to their complex structures, making them unsuitable for IoT devices. In this paper, a lightweight and efficient NIDS named DFE is proposed, using a neural network based on deep feature extraction. The DFE model extracts highly discriminative features with a small number of layers and calculations, making it ideal for real-time intrusion detection in IoT devices with limited processing capabilities.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Chemistry, Analytical
Safi Ullah, Jawad Ahmad, Muazzam A. Khan, Eman H. Alkhammash, Myriam Hadjouni, Yazeed Yasin Ghadi, Faisal Saeed, Nikolaos Pitropakis
Summary: This paper proposes a deep-convolutional-neural-network-based intrusion detection system to improve performance and reduce computational power. Experiments were conducted using the IoTID20 dataset, and performance analysis was carried out with metrics such as accuracy, precision, recall, and F1-score.
Article
Computer Science, Information Systems
Nabil Sabor, Mohamed Abdelraheem
Summary: In this paper, a efficient crowdsensing system named CMNN-RADC was proposed for detecting and classifying road anomalies. The system utilizes low and high frequency features extracted from three acceleration signals (x, y, and z) and combines them using a two-stages Convolutional-based Mixer Neural Network. Experimental results show that CMNN-RADC outperforms previous methods, achieving a detection rate of 99.3% and classification F1 scores of 98.42%, 97.68%, and 99.05% for metal bump, pothole, and asphalt bump, respectively.
INTERNET OF THINGS
(2023)
Article
Computer Science, Information Systems
Tanzeela Altaf, Xu Wang, Wei Ni, Guangsheng Yu, Ren Ping Liu, Robin Braun
Summary: In recent years, there has been an increase in sophisticated cyber-attacks causing financial instability and privacy breaches. This highlights the need for a more robust and effective Network Intrusion Detection System (NIDS) that can learn complex relations of network flows. The proposed GNN-based NIDS utilizes a novel graph structure to capture full communication between IoT nodes, achieving significant improvements in accuracy, precision, F1, and recall compared to state-of-the-art models.
INTERNET OF THINGS
(2023)
Article
Chemistry, Analytical
Basim Ahmad Alabsi, Mohammed Anbar, Shaza Dawood Ahmed Rihan
Summary: This paper presents an approach for detecting attacks on IoT networks using a combination of two convolutional neural networks (CNN-CNN). The results show that the proposed approach achieves high accuracy, precision, recall, and classification rate, and outperforms other deep learning algorithms and feature selection methods.
Article
Computer Science, Artificial Intelligence
David A. Bierbrauer, Michael J. De Lucia, Krishna Reddy, Paul Maxwell, Nathaniel D. Bastian
Summary: Traditional machine learning models for network intrusion detection rely on extensive network traffic data, which is not feasible in edge network scenarios. This study explores the feasibility of using transfer learning and neural networks to perform intrusion detection solely based on raw network traffic in computationally limited environments. The results show that high accuracy can be achieved on edge devices with a combination of transferred one-dimensional convolutional neural network model and retrained random forest model.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Petroleum
Xiaopeng Ma, Kai Zhang, Jian Wang, Chuanjin Yao, Yongfei Yang, Hai Sun, Jun Yao
Summary: This paper introduces a deep learning-based surrogate modeling framework that can directly predict production data from high-dimensional spatial parameters. By combining this surrogate model with an improved data assimilation algorithm, a surrogate-based history-matching workflow is developed.
Article
Computer Science, Information Systems
Tim M. Booij, Irina Chiscop, Erik Meeuwissen, Nour Moustafa, Frank T. H. den Hartog
Summary: This article introduces a novel IoT data set, ToN_IoT, and provides its description, statistical analysis, and machine learning evaluation. Comparisons to other data sets show the importance of heterogeneity and the impact of differences on detection performance. It is found that including different data collection methods and diverse monitored features are crucial for the usefulness of IoT network intrusion data sets. Practical application requires standardization of feature descriptions and cyberattack classes, requiring a joint effort from the research community.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Yanbin Hao, Shuo Wang, Yi Tan, Xiangnan He, Zhenguang Liu, Meng Wang
Summary: Efficient action recognition is achieved through a novel spatio-temporal collaborative (STC) module, which integrates channel splitting and filter decoupling for efficient architecture design and feature refinement. Experimental results demonstrate that the proposed STC networks strike a competitive balance between model efficiency and effectiveness in video action recognition tasks.
IEEE TRANSACTIONS ON IMAGE PROCESSING
(2022)
Article
Computer Science, Information Systems
Bahman Varastan, Shahram Jamali, Reza Fotohi
Summary: The Internet of Things (IoTs) is a complex network covering various areas, and protecting these networks from attacks is of particular importance. This study proposes a new and effective detection method based on LSTM-RNN and PCA.
CLUSTER COMPUTING-THE JOURNAL OF NETWORKS SOFTWARE TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
N. Krishnavardhan, M. Govindarajan, S. V. Achutha Rao
Summary: With the increasing use of the internet, more and more industries, including finance, are providing services online. Financial fraud has become a major issue globally, causing significant financial losses. Machine learning and data mining approaches have been extensively used to address this problem, but further development is needed to handle massive data and identify new attack patterns quickly. In this research, a deep learning-based approach using the stacked temporal convolution network technique is proposed to improve the efficiency and accuracy of financial fraud detection. A flower pollination optimization process is also incorporated for feature selection to address potential side effects. The suggested FPO_QCNN achieves high accuracy for credit card, insurance, and mortgage fraud detection according to the experimental results.
Article
Computer Science, Information Systems
Jiawei Du, Kai Yang, Yanjing Hu, Lingjie Jiang
Summary: This paper constructs a network intrusion detection classification model (NIDS-CNNLSTM) based on deep learning for the wireless sensing scenario of the Industrial Internet of Things (IIoT) to effectively distinguish and identify network traffic data and ensure the security of the equipment and operation of the IIoT.
Article
Engineering, Civil
Yan Chen, Tian Shu, Xiaokang Zhou, Xuzhe Zheng, Akira Kawai, Kaoru Fueda, Zheng Yan, Wei Liang, Kevin I-Kai Wang
Summary: In this paper, a Graph Attention Network with Spatial-Temporal Clustering (GAT-STC) is proposed to improve traffic flow forecasting in Intelligent Transportation System (ITS). The network extracts recent-aware features and periodic-aware features to capture dynamic changes in spatial feature representation. Experimental results show that the proposed model outperforms five baseline methods in terms of accuracy and efficiency.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Information Systems
Xiaokang Zhou, Wei Liang, Weimin Li, Ke Yan, Shohei Shimizu, Kevin I-Kai Wang
Summary: The study introduces a novel adversarial attack generation method to degrade the classification precision of intelligent intrusion detection in IoT systems by identifying critical feature elements and minimal perturbations. The method also develops a hierarchical node selection algorithm based on random walk with restart to select more vulnerable nodes.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Automation & Control Systems
Xiaohan Tu, Cheng Xu, Siping Liu, Renfa Li, Guoqi Xie, Jing Huang, Laurence Tianruo Yang
Summary: This article addresses the shortcomings of monocular depth estimation (MDE) methods by designing an efficient model and utilizing a reinforcement learning algorithm for automatic channel pruning. Through pruning and compilation optimization, experimental results demonstrate the effectiveness of our methods in achieving accurate depth sensing on different hardware architectures.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Review
Medical Informatics
Amine Boulemtafes, Abdelouahid Derhab, Yacine Challal
Summary: This paper focuses on the privacy preservation issue in pervasive health monitoring applications, especially in constrained client-side environments. It reviews the adequacy of existing privacy-preserving solutions and discusses evaluation criteria and future research directions.
HEALTH AND TECHNOLOGY
(2022)
Article
Computer Science, Hardware & Architecture
Alaeddine Mihoub, Ouissem Ben Fredj, Omar Cheikhrouhou, Abdelouahid Derhab, Moez Krichen
Summary: This paper investigates the detection of DoS/DDoS attacks in IoT using machine learning techniques. A new architecture is proposed, consisting of two components: DoS/DDoS detection and mitigation. The detection component provides fine-granularity detection, identifying the specific type of attack and packet type used. Evaluation on the Bot-IoT dataset shows promising results, with a Looking-Back-enabled Random Forest classifier achieving an accuracy of 99.81%.
COMPUTERS & ELECTRICAL ENGINEERING
(2022)
Article
Engineering, Civil
Abdelouahid Derhab, Mohamed Belaoued, Irfan Mohiuddin, Fajri Kurniawan, Muhammad Khurram Khan
Summary: In this paper, a Histogram-based Intrusion Detection and Filtering framework called H-IDFS is proposed, which assembles CAN packets into windows and computes histograms for classification. A novel one-class SVM named OCSVM-attack is introduced for filtering out normal CAN packets from malicious windows. Experimental results demonstrate the superiority of H-IDFS in window classification and normal packet filtering.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Engineering, Chemical
Fahad R. Albogamy, Yasir Ashfaq, Ghulam Hafeez, Sadia Murawwat, Sheraz Khan, Faheem Ali, Farrukh Aslam Khan, Khalid Rehman
Summary: This study proposes a framework for demand-side management (DSM) by scheduling energy consumption using a flat pricing scheme (FPS) in a smart grid (SG). The framework includes a microgrid with renewable energy sources, energy storage systems, electric vehicles (EVs), and building appliances. The ant colony optimization (ACO) algorithm efficiently schedules smart appliances and EVs batteries charging/discharging to minimize energy cost, carbon emission, and peak to average ratio (PAR). An integrated technique of enhanced differential evolution (EDE) algorithm and artificial neural network (ANN) is used for accurate microgrid energy estimation. Simulations are conducted to test the applicability of the proposed framework and compare it to other scheduling energy management frameworks. The results show significant reductions in energy cost, PAR, and carbon emission compared to the non-scheduling case, affirming the effectiveness of the proposed framework.
Article
Engineering, Chemical
Sana Ullah, Ghulam Hafeez, Gul Rukh, Fahad R. Albogamy, Sadia Murawwat, Faheem Ali, Farrukh Aslam Khan, Sheraz Khan, Khalid Rehman
Summary: Agricultural productivity is crucial for a country's economy, and the proper provision of water is essential for increasing productivity. A smart-sensors-based solar-powered system has been developed to monitor and control water supply to crops, reducing water and energy wastage.
Article
Energy & Fuels
Sajawal Ur Rehman Khan, Israa Adil Hayder, Muhammad Asif Habib, Mudassar Ahmad, Syed Muhammad Mohsin, Farrukh Aslam Khan, Kainat Mustafa
Summary: Nowadays, electric load forecasting through a data analytic approach has become one of the most active and emerging research areas. It provides future consumption patterns of electric load. By analyzing past electric consumption records to estimate the upcoming electricity load, the issue of fluctuating behavior can be resolved. In this study, a framework for feature selection, extraction, and regression is proposed to carry out the electric load prediction.
Article
Telecommunications
Abdelouahid Derhab, Omar Cheikhrouhou, Azza Allouch, Anis Koubaa, Basit Qureshi, Mohamed Amine Ferrag, Leandros Maglaras, Farrukh Aslam Khan
Summary: Drones have become a significant technological breakthrough, especially with the integration of the Internet, forming the Internet of Drones (IoD). This paper provides a comprehensive survey on the cyber and physical security of IoD, including classifications of assets, attacks, and countermeasures. It introduces the concept of Chain of Impact to evaluate the risk of attacks and proposes a taxonomy of countermeasures. The paper also identifies research challenges and suggests future directions for IoD security. Rating: 8/10
VEHICULAR COMMUNICATIONS
(2023)
Article
Energy & Fuels
Ahmad Alzahrani, Khizar Sajjad, Ghulam Hafeez, Sadia Murawwat, Sheraz Khan, Farrukh Aslam Khan
Summary: Real-time energy optimization is crucial for load scheduling, cost reduction, demand and supply balance, and power system reliability. However, the unpredictable nature of renewable energy and electric loads poses challenges for real-time optimization. The Lyapunov optimization technique has emerged as a solution to this problem. This research investigates a smart home with various loads and renewable energy sources in a grid-connected mode to optimize cost and energy storage using the Lyapunov optimization technique.
Article
Chemistry, Analytical
Shayan E. Ali, Noshina Tariq, Farrukh Aslam Khan, Muhammad Ashraf, Wadood Abdul, Kashif Saleem
Summary: Sensitive applications like healthcare and medical services require reliable transmission for the success of communication technology. However, these systems are vulnerable to attacks like Sybil, where false nodes are created. To address this, a blockchain-based fuzzy trust management framework (BFT-IoMT) is proposed to detect and isolate Sybil nodes in healthcare applications. The results show that BFT-IoMT is more efficient and effective compared to other frameworks in terms of energy consumption, attack detection, trust computation reliability, packet delivery ratio, and throughput.
Article
Information Science & Library Science
Ibrahim Mutambik, Abdullah Almuqrin, Yulong David Liu, Waleed Halboob, Abdullah Alakeel, Abdelouahid Derhab
Summary: Many countries are implementing open government data (OGD) initiatives, but these initiatives often fail to attract continuous use and deliver a satisfactory return on investment. A study identified four factors that strongly influence the intention to use OGD, which can be helpful for policymakers to formulate strategies that drive up continuous OGD engagement.
JOURNAL OF GLOBAL INFORMATION MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
ElMouatez Billah Karbab, Mourad Debbabi, Abdelouahid Derhab
Summary: To address the issue of ransomware attacks, we propose a new portable framework called SwiftR, which can perform cross-platform ransomware detection and fingerprinting. SwiftR uses advanced deep learning techniques and special feature extraction methods to accurately detect ransomware in both static and dynamic analysis. Extensive evaluation shows that SwiftR achieves excellent performance in ransomware detection, segregation, and family attribution.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Amine Boulemtafes, Abdelouahid Derhab, Yacine Challal
Summary: The remote deep learning paradigm is suitable for pervasive health monitoring (PHM) applications as it addresses the constraints of client-side environments. However, ensuring high accuracy, client-side constraints, and privacy requirements related to health data sensitivity remain challenging. In this study, a privacy-preserving remote inference solution called PRIviLY is proposed for Fully Connected Deep Networks (FCDNs), which achieves significant improvements in communication and computation overhead without compromising accuracy.
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Faiza Babar Khan, Muhammad Hanif Durad, Asifullah Khan, Farrukh Aslam Khan, Sajjad Hussain Chauhdary, Mohammed Alqarni
Summary: Malware is a significant threat to information security, and efficient anti-malware software is crucial for protection. However, identifying malware remains challenging, especially with unknown samples. In this paper, a novel architecture based on the Relation Network is proposed for Few-Shot Learning (FSL) implementation, achieving improved classification accuracy by up to 94% with only one training instance.
Article
Computer Science, Information Systems
Muhammad Hassan, Noshina Tariq, Amjad Alsirhani, Abdullah Alomari, Farrukh Aslam Khan, Mohammed Mujib Alshahrani, Muhammad Ashraf, Mamoona Humayun
Summary: In this paper, a novel fog-enabled GINI index-based trust mechanism (GITM) is proposed to mitigate Sybil attacks in the smart grid. GITM detects and isolates a greater number of malicious network nodes compared to other techniques within a similar time frame by utilizing the forwarding behavior of legitimate member nodes. By using the proposed GITM framework, the Sybil attack detection rate increases by 4.48%, energy consumption reduces by 21%, and isolation latency reduces by 26.30% (concerning time). Furthermore, the end-to-end delay is merely 0.30% more in our case, and the number of control messages decreases by 28%.
Article
Computer Science, Cybernetics
Abdelouahab Amira, Abdelouahid Derhab, Samir Hadjar, Mustapha Merazka, Md. Golam Rabiul Alam, Mohammad Mehedi Hassan
Summary: The widespread use of social media platforms has led to an increase in the dissemination of fake news. This study focuses on detecting organized groups participating in fake news campaigns without prior knowledge of the news content or social account profiles. A spatial-temporal similarity graph is proposed to connect social accounts involved in similar fake news campaigns. A community detection algorithm is applied to cluster users into communities, and a labeling algorithm is used to label communities as benign or malicious based on a fake news classifier. Evaluation results show high accuracy in community labeling, and statistical analysis identifies significant structural features between benign and malicious communities.
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS
(2023)