SAFORI‐IDS: Semantic‐Aware Federated Intrusion Detection for IoT‐Enabled 6G Metaverse Infrastructure
TL;DR
The original SAFORI‐IDS framework uniquely merges semantic context and federated privacy to secure 6G‐enabled Metaverse environments and achieves over 99% accuracy, precision, recall, and F1‐score, significantly outperforming state‐of‐the‐art baselines while maintaining lightweight computational efficiency suitable for IoT edge devices.
Abstract
The proliferation of Internet of things (IoT) devices and the emerging 6G Metaverse introduces unprecedented opportunities, but also exposes massive attack surfaces risking severe privacy breaches and physical dangers. Conventional intrusion detection systems (IDS), whether rule‐based or deep learning‐based, fail to mitigate these critical threats under heterogeneous non‐IID traffic, adversarial perturbations, and stringent resource constraints. To address these challenges, our original SAFORI‐IDS framework uniquely merges semantic context and federated privacy to secure 6G‐enabled Metaverse environments. The framework integrates multimodal feature extraction, semantic representation learning, and federated aggregation to ensure privacy‐preserving and scalable detection across distributed clients. An explainability module based on SHAP enhances interpretability by attributing detection outcomes to meaningful features, improving trustworthiness for security administrators. Experimental evaluation on the CICIDS2017 dataset demonstrates that SAFORI‐IDS achieves over 99% accuracy, precision, recall, and F1‐score, significantly outperforming state‐of‐the‐art baselines while maintaining lightweight computational efficiency suitable for IoT edge devices.
