IoE-GraphFormer: A Graph Transformer-Based Framework for Anomaly Detection in Internet of Everything

The Internet of Everything (IoE) integrates people, processes, data, and things into highly interconnected, heterogeneous, and dynamic networks, significantly increasing exposure to cybersecurity threats. Traditional cybersecurity mechanisms designed for the Internet of Things (IoT) often fall short when faced with IoE's inherent complexity and scale. This paper presents IoE-GraphFormer, a novel cybersecurity framework leveraging advanced graph transformer architectures to effectively detect anomalies and cyber threats in complex IoE networks. IoEGraphFormer models intricate relationships among diverse entities within IoE environments, capturing long-range dependencies and dynamic interactions critical to accurate anomaly detection. Experiments performed on simulated IoE datasets demonstrate superior detection accuracy, robustness, and scalability compared to existing graph-based cybersecurity solutions, underscoring the potential of graph transformers to safeguard next-generation IoE systems