Federated Temporal RAG for Privacy-Preserving Real-Time Threat Intelligence in Distributed Remote Patient Monitoring

This paper presents FedTempRAG, a federated and privacy-preserving framework for real-time threat intelligence in distributed Remote Patient Monitoring (RPM) systems. By integrating decentralized retrieval, temporal knowledge graphs, and secure aggregation, the proposed approach enables collaborative attack detection without sharing sensitive data. Experimental results on multi-hospital IoMT environments demonstrate high detection accuracy, low latency, and strong resilience against adversarial behavior, highlighting its effectiveness for secure and practical healthcare deployments.