BlockFed: Blockchain-based Privacy Preserving Federated Learning for 5G-assisted Healthcare Ecosystems

With the rapid adoption of 5G networks and the growing reliance on digital healthcare, the need for secure, efficient, and privacy-aware data processing has become increasingly critical. This paper presents a novel approach BlockFed , that integrates Federated Learning (FL) with Blockchain (BC) technology to ensure data privacy and model integrity in 5G-assisted healthcare ecosystems. In the proposed system, Patient Health Record (PHR) remains at local Healthcare Entities (HE) such as hospitals and research centers, and only encrypted model updates are shared, effectively preserving user privacy. BC is employed to record and verify model weight transactions, providing tamper-proof integrity and transparency among participating HE. To mitigate the high storage demands of BC, the InterPlanetary File System (IPFS) is utilized for off-chain storage of model weights. Additionally, a lightweight homomorphic encryption scheme is incorporated to protect model parameters during aggregation and transmission. This integrated approach offers a scalable and trustworthy solution for collaborative healthcare intelligence while safeguarding sensitive PHR. Experimental insights and theoretical validation demonstrate the system’s potential for practical deployment in next-generation healthcare infrastructures.