Personalized Federated Learning based Intrusion Detection Approach for Metaverse
Gurpreet Singh, P. Rajalakshmi, Yong Xiang · 2023
Metaverse is an emerging paradigm aiming to offer immersive 3D experiences and self-sustaining virtual shared environments. Thanks to the rapid progress in technologies like 5G, Augmented Reality (AR), Virtual Reality (VR), and mixed reality in real-time applications, the Metaverse is evolving from science fiction to extended reality (XR). Next-generation networks offer new services for example Metaverse but also expose vulnerabilities, making them susceptible to exploitation by threat actors. It is proved that privacy is a big concern in Metaverse posing risks to data confidentiality, integrity, and availability. With the growing importance of data privacy and security, Federated Learning (FL) is a promising technology. This enables the training of machine learning models using decentralized data, preserving privacy by design. To safeguard user data privacy, we investigate the possibility of using a Personalized Federated Learning-based Intrusion Detection System (pFed-IDS). We employed the latest datasets to assess model performance in non-IID scenarios. Our early work provides insights to secure user data in next-generation technologies.