The rapid adoption of Internet of Things (IoT) devices in healthcare has introduced new challenges in preserving data privacy, security and patient safety. Traditional approaches need to ensure security and privacy while maintaining computational efficiency, particularly for resource-constrained IoT devices. This paper proposes a novel hybrid approach combining federated learning and blockchain technology to provide a secure and privacy-preserved solution for IoT-enabled healthcare applications. Our approach leverages a public-key cryptosystem that provides semantic security for local model updates, while blockchain technology ensures the integrity of these updates and enforces access control and accountability. The federated learning process enables a secure model aggregation without sharing sensitive patient data. We implement and evaluate our proposed framework using EMNIST datasets, demonstrating its effectiveness in preserving data privacy and security while maintaining computational efficiency. The results suggest that our hybrid approach can significantly enhance the development of secure and privacy-preserved IoT-enabled healthcare applications, offering a promising direction for future research in this field.

翻译：物联网设备在医疗领域的快速普及带来了数据隐私、安全性及患者安全方面的新挑战。传统方法在保障安全性和隐私性的同时，需要兼顾计算效率，尤其对于资源受限的物联网设备而言更具难度。本文提出了一种融合联邦学习与区块链技术的新型混合方法，为物联网医疗应用提供安全且保护隐私的解决方案。该方法采用公钥密码系统为本地模型更新提供语义安全性，同时借助区块链技术确保更新的完整性并实施访问控制与问责机制。联邦学习过程通过不共享敏感患者数据实现安全模型聚合。我们利用EMNIST数据集对所提框架进行了实现与评估，验证了其在维持计算效率的同时有效保护数据隐私与安全性的能力。结果表明，该混合方法能显著提升物联网医疗应用的安全性与隐私保护水平，为相关领域未来研究提供了有前景的方向。