Intrusion detection in metaverse environment internet of things systems by metaheuristics tuned two level framework

Internet of Things (IoT) is one of the most important emerging technologies that supports Metaverse integrating process, by enabling smooth data transfer among physical and virtual domains. Integrating sensor devices, wearables, and smart gadgets into Metaverse environment enables IoT to deepen inte...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Scientific reports Ročník 15; číslo 1; s. 3555 - 31
Hlavní autoři: Antonijevic, Milos, Zivkovic, Miodrag, Djuric Jovicic, Milica, Nikolic, Bosko, Perisic, Jasmina, Milovanovic, Marina, Jovanovic, Luka, Abdel-Salam, Mahmoud, Bacanin, Nebojsa
Médium: Journal Article
Jazyk:angličtina
Vydáno: London Nature Publishing Group UK 28.01.2025
Nature Publishing Group
Nature Portfolio
Témata:
631/114/2164
631/114/2397
631/114/2400
631/114/2401
Artificial intelligence
CatBoost
Chimp optimization algorithm
Classification
Data collection
Datasets
Decision making
Electrical engineering
Humanities and Social Sciences
Internet
Internet of Things
LightGBM
Machine learning
Metaheuristics algorithms
Metaverse
multidisciplinary
Neural networks
Optimization
Optimization algorithms
Science
Science (multidisciplinary)
Metaheuristics algorithms
Chimp optimization algorithm
Metaverse
LightGBM
CatBoost
Optimization
ISSN:2045-2322, 2045-2322
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Internet of Things (IoT) is one of the most important emerging technologies that supports Metaverse integrating process, by enabling smooth data transfer among physical and virtual domains. Integrating sensor devices, wearables, and smart gadgets into Metaverse environment enables IoT to deepen interactions and enhance immersion, both crucial for a completely integrated, data-driven Metaverse. Nevertheless, because IoT devices are often built with minimal hardware and are connected to the Internet, they are highly susceptible to different types of cyberattacks, presenting a significant security problem for maintaining a secure infrastructure. Conventional security techniques have difficulty countering these evolving threats, highlighting the need for adaptive solutions powered by artificial intelligence (AI). This work seeks to improve trust and security in IoT edge devices integrated in to the Metaverse. This study revolves around hybrid framework that combines convolutional neural networks (CNN) and machine learning (ML) classifying models, like categorical boosting (CatBoost) and light gradient-boosting machine (LightGBM), further optimized through metaheuristics optimizers for leveraged performance. A two-leveled architecture was designed to manage intricate data, enabling the detection and classification of attacks within IoT networks. A thorough analysis utilizing a real-world IoT network attacks dataset validates the proposed architecture’s efficacy in identification of the specific variants of malevolent assaults, that is a classic multi-class classification challenge. Three experiments were executed utilizing data open to public, where the top models attained a supreme accuracy of 99.83% for multi-class classification. Additionally, explainable AI methods offered valuable supplementary insights into the model’s decision-making process, supporting future data collection efforts and enhancing security of these systems.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2045-2322
2045-2322
DOI:10.1038/s41598-025-88135-9