A Stable Generative Adversarial Network Architecture for Network Intrusion Detection

Soleymanzadeh, Raha; Kashef, Rasha

doi:10.1109/csr54599.2022.9850286

Cited by 7 publications

(3 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study in [21] introduced an ensemble deep learning technique for the classification of network attacks. The authors built a robust generative adversarial network based on ensemble convolutional neural networks (GANsECNN).…”

Section: Deep Learning Methodsmentioning

confidence: 99%

“…This challenge points to an inherent limitation in the current autoencoder architectures used for cybersecurity purposes. Furthermore, while methods like those proposed by Raha et al [21] for the generation of synthetic data to train robust models show promise, they often do not outperform existing benchmarks, indicating a gap in the efficacy of such generative approaches. Similarly, the use of federated learning in medical cyber-physical systems, as discussed in Ilias et al [22], although improving the performance over non-federated models, still shows sub-optimal results in certain key areas like network flow anomaly detection.…”

Section: Deep Learning Methodsmentioning

confidence: 99%

“…They can be applied to more sophisticated tasks like image processing and natural language processing. The following state-of-the-art works have adopted DL (non-Transformer-based) methods: [20][21][22][25][26][27][28][29]. The following works have combined DL with ML methods: [30][31][32][33][34][35].…”

Section: Other Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Navigating the Cyber Threat Landscape: An In-Depth Analysis of Attack Detection within IoT Ecosystems

AboulEla,

Ibrahim,

Shehmir

et al. 2024

Self Cite

View full text Add to dashboard Cite

The Internet of Things (IoT) is seeing significant growth, as the quantity of interconnected devices in communication networks is on the rise. The increased connectivity of devices has heightened their susceptibility to hackers, underscoring the need to safeguard IoT devices. This research investigates cybersecurity in the context of the Internet of Medical Things (IoMT), which encompasses the cybersecurity mechanisms used for various healthcare devices connected to the system. This study seeks to provide a concise overview of several artificial intelligence (AI)-based methodologies and techniques, as well as examining the associated solution approaches used in cybersecurity for healthcare systems. The analyzed methodologies are further categorized into four groups: machine learning (ML) techniques, deep learning (DL) techniques, a combination of ML and DL techniques, Transformer-based techniques, and other state-of-the-art techniques, including graph-based methods and blockchain methods. In addition, this article presents a detailed description of the benchmark datasets that are recommended for use in intrusion detection systems (IDS) for both IoT and IoMT networks. Moreover, a detailed description of the primary evaluation metrics used in the analysis of the discussed models is provided. Ultimately, this study thoroughly examines and analyzes the features and practicality of several cybersecurity models, while also emphasizing recent research directions.

show abstract

Section: Deep Learning Methodsmentioning

confidence: 99%

Section: Deep Learning Methodsmentioning

confidence: 99%