Related references
Note: Only part of the references are listed.
Article
Automation & Control Systems
Pedro Ruzafa-Alcazar et al.
Summary: Federated learning (FL) has gained significant attention for its advantages and applicability, but sharing updated gradients/weights in the training process raises privacy concerns, especially in the context of IoT. Our work comprehensively evaluates differential privacy techniques applied to an FL-enabled intrusion detection system (IDS) for industrial IoT, considering nonindependent and identically distributed data. We compare the accuracy achieved with different privacy requirements and aggregation functions (FedAvg and Fed+), and find that using Fed+ yields similar results even with noise in the federated training process.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Computer Science, Information Systems
Rahul Yumlembam et al.
Summary: This study demonstrates the effectiveness of graph-based deep learning for detecting malicious Android apps and proposes a generative adversarial network algorithm to attack this detection method. Experimental analysis shows that the proposed algorithm can effectively reduce the detection rate of malicious apps and retraining the model helps combat adversarial attacks.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Information Systems
Pakarat Musikawan et al.
Summary: Android-based mobile devices have become popular due to their ease of use and wide range of capabilities. However, this popularity has also attracted attackers who use sophisticated malware obfuscation and detection avoidance tactics. In this article, an improved deep neural network called AMDI-Droid is presented to safeguard Android devices from malicious apps. The model combines multiple hidden layers to learn effective feature representations and uses a blending approach to produce final predictions.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Cybernetics
Vinayakumar Ravi et al.
Summary: This study proposes an attention-based multidimensional deep learning approach for cross-architecture IoMT malware detection, classification, and CPU architectures classification. Experimental results show that the proposed method achieves high accuracy and performs well on multiple datasets.
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS
(2023)
Article
Computer Science, Hardware & Architecture
Xinjun Pei et al.
Summary: This paper proposes a semi-supervised federated IoT malware detection framework named FedMalDE, which uses knowledge transfer technologies to infer labels of unlabeled samples based on the correlation between labeled and unlabeled data. It also employs a specially designed subgraph aggregated capsule network (SACN) to capture varied malicious behaviors efficiently. The experiments demonstrate the effectiveness of FedMalDE in detecting IoT malware and its sufficient privacy and robustness guarantee.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2023)
Article
Computer Science, Hardware & Architecture
Valerian Rey et al.
Summary: With the increasing number of IoT devices and the growing importance of data privacy and security, researching the application and security issues of federated learning in IoT malware detection becomes crucial. This study explores the use of federated learning to detect malware while preserving data privacy and finds that it has the capability to detect malware, but further efforts are needed to enhance its robustness.
Article
Computer Science, Information Systems
Viraaji Mothukuri et al.
Summary: The Internet of Things (IoT) consists of billions of physical devices connected to the Internet, performing tasks independently with less human intervention. However, IoT networks are vulnerable to malicious attacks that aim to steal and manipulate personal data. In order to address this issue, the paper proposes a federated-learning (FL)-based approach that uses decentralized on-device data for anomaly detection in IoT networks. Experimental results demonstrate that this approach outperforms traditional centralized machine learning methods in securing user data privacy and achieving optimal accuracy in attack detection.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Automation & Control Systems
Mamoun Alazab et al.
Summary: This article presents the application of federated learning in enhancing cybersecurity and preventing cyberattacks in real-time scenarios. The authors conducted a comprehensive survey on various federated learning models developed by researchers for authentication, privacy, trust management, and attack detection. Real-time use cases and the adoption of federated learning for data privacy and system performance improvement are also discussed. The article concludes with prominent challenges and future directions for researchers to focus on.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Baoguo Yuan et al.
Summary: This article proposes an IoT malware classification method based on lightweight convolutional neural networks (LCNNs). The method accurately identifies and classifies malware variants in IoT systems, achieving high accuracy while reducing model size.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Physics, Multidisciplinary
Changnan Jiang et al.
Summary: With the increasing popularity of Android and the risks posed by hackers, the detection and classification of malware on the Android platform has become a research focus. However, existing methods rely on complex manual operations or large amounts of high-quality training data, while the collected malware data contains user privacy information. This study proposes a new Android malware classification scheme based on Federated learning, which can protect user privacy and achieve higher accuracy compared to existing methods.
Article
Automation & Control Systems
Bushra Tahir et al.
Summary: This article introduces a method for detecting false data injection attacks in an Internet-of-Things-based transactive energy system. By utilizing recurrent deep deterministic policy gradient, a solution is designed to detect such attacks in a complex environment. Additionally, a decentralized FDIA detection method based on deep federated learning is employed, using attentive aggregation to enable independent clients to train a centralized model while maintaining data privacy intact.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Automation & Control Systems
Tu N. Nguyen et al.
Summary: In recent years, attackers have increasingly targeted IoT devices in the industrial Internet of things (IIoT), with IoT botnet emerging as the most urgent security issue. The main methods for detecting IoT botnets are static, dynamic, and hybrid analysis. This article presents a novel method using dynamic analysis to enhance graph-based features generated from static analysis for IoT botnet detection. Experimental results show that this approach outperforms existing methods in terms of accuracy and complexity.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Information Systems
Jinhui Ning et al.
Summary: This article proposes three methods based on semi-supervised learning, transfer learning, and domain adaptive to improve the accuracy of malware traffic classification in Industrial Internet of Things using a large amount of unlabeled data. Experimental results demonstrate the satisfactory performance of these methods in the case of few labeled samples.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Proceedings Paper
Computer Science, Information Systems
Scott Freitas et al.
Summary: Computer vision plays a crucial role in automated malware detection, but current research is limited by the availability of datasets. MalNetImage, the largest public cybersecurity image database, provides more images and classes, unlocking new opportunities for advancing machine learning and exploring new research directions.
PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2022
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Liangqiong Qu et al.
Summary: This study demonstrates that self-attention-based architectures, such as Transformers, are more robust for federated learning with heterogeneous data. They reduce forgetting, accelerate convergence, and improve the global model.
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2022)
Article
Computer Science, Information Systems
Sachith Seneviratne et al.
Summary: This paper presents a novel deep learning model called SHERLOCK based on self-supervised learning for malware detection, which achieves high accuracy in binary classification and multi-class classification of malware according to experimental results.
Article
Computer Science, Information Systems
Zahra Moti et al.
Summary: This paper proposes a new method for detecting and generating malware, called MalGan. By combining CNN and Generative Adversarial Network techniques, a significant number of previously unseen malware samples can be detected with high accuracy even with only a few malware samples.
Article
Automation & Control Systems
Rahim Taheri et al.
Summary: Fed-IIoT is a robust architecture for detecting Android malware applications in IIoT, utilizing federated learning on both participant and server sides to enhance security and privacy protection.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Computer Science, Information Systems
Qi Li et al.
Summary: This article proposes a method for detecting malware variants in the IoT, using techniques such as feature representation, convolutional neural networks to address the differences in malicious code between traditional networks and IoT platforms, effectively detecting malware variants in the IoT.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Computer Science, Information Systems
Naercio Magaia et al.
Summary: The significant evolution of the Internet of Things has led to the development of smart city devices that have replaced manual labor, increasing efficiency and intelligence in cities. However, the increased sensitivity of data, especially in the industrial sector, has attracted hackers targeting Industrial IoT devices or networks, leading to a rise in the number of malware infections. This article discusses the concept and applications of IIoT in smart cities, as well as the security challenges faced in this emerging area, along with available deep learning techniques to enhance IIoT security.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Biochemical Research Methods
Pauli Virtanen et al.
Article
Computer Science, Information Systems
Mohammed Ali Al-Garadi et al.
IEEE COMMUNICATIONS SURVEYS AND TUTORIALS
(2020)
Review
Computer Science, Information Systems
Omer Aslan et al.
