A Comprehensive Privacy-Preserving Federated Learning Scheme With Secure Authentication and Aggregation for Internet...

Area of Science:

• Computer Science
• Medical Informatics
• Cybersecurity

Background:

• The Internet of Medical Things (IoMT) generates vast amounts of data, necessitating advanced data mining and integration.
• Federated learning offers a privacy-preserving approach for training models on distributed medical data.
• Existing federated learning methods are vulnerable to privacy leakage and security threats, especially in healthcare.

Purpose of the Study:

• To propose a comprehensive privacy-preserving federated learning scheme tailored for IoMT systems.
• To address security vulnerabilities and privacy leakage risks inherent in current federated learning implementations.
• To enhance the security and efficiency of federated learning in resource-constrained IoMT environments.

Main Methods:

• Implementation of a novel federated learning scheme incorporating blind masking.
• Utilization of certificateless proxy re-encryption (CL-PRE) for secure model aggregation.
• Integration of a tactful dropout handling mechanism to improve robustness.

Main Results:

• The proposed scheme ensures the confidentiality of local medical data during federated training.
• Achieved secure aggregation of local models, protecting against privacy leakage.
• Provided robust authentication for uploaded models while preserving user identity privacy.
• Demonstrated superior performance in functional features and efficiency compared to existing schemes.

Conclusions:

• The developed privacy-preserving federated learning scheme effectively safeguards sensitive medical data in IoMT.
• The scheme offers enhanced security, privacy, and efficiency, making it highly applicable to large-scale IoMT deployments.
• This approach mitigates risks associated with open wireless channels and malicious attacks in IoMT.