SMPC-based Federated Learning for 6G enabled Internet of Medical Things

Aditya Pribadi Kalapaaking,Veronika Stephanie,Ibrahim Khalil,Mohammed Atiquzzaman,Xun Yi,Mahathir Almashor

DOI: https://doi.org/10.1109/MNET.007.2100717

2023-04-26

Abstract:Rapidly developing intelligent healthcare systems are underpinned by Sixth Generation (6G) connectivity, ubiquitous Internet of Things (IoT), and Deep Learning (DL) techniques. This portends a future where 6G powers the Internet of Medical Things (IoMT) with seamless, large-scale, and real-time connectivity amongst entities. This article proposes a Convolutional Neural Network (CNN) based Federated Learning framework that combines Secure Multi-Party Computation (SMPC) based aggregation and Encrypted Inference methods, all within the context of 6G and IoMT. We consider multiple hospitals with clusters of mixed IoMT and edge devices that encrypt locally trained models. Subsequently, each hospital sends the encrypted local models for SMPC-based encrypted aggregation in the cloud, which generates the encrypted global model. Ultimately, the encrypted global model is returned to each edge server for more localized training, further improving model accuracy. Moreover, hospitals can perform encrypted inference on their edge servers or the cloud while maintaining data and model privacy. Multiple experiments were conducted with varying CNN models and datasets to evaluate the proposed framework's performance.

Cryptography and Security

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to address the issue of implementing a privacy-preserving federated learning framework in the Internet of Medical Things (IoMT) environment supported by sixth-generation communication technology (6G) through secure multi-party computation (SMPC). Specifically, the goals of the paper include:

1. **Ensuring the privacy of local models**: During the federated learning process among multiple hospitals, ensuring that each hospital's local model is not leaked or tampered with during aggregation.
2. **Encrypting the generation and inference of the global model**: Proposing a secure encryption method based on SMPC, so that the global model is protected not only during its generation but also maintains data and model privacy when performing inference on a third-party cloud platform.
3. **Improving the accuracy of medical diagnosis**: Enhancing the accuracy and coverage of diagnoses through collaborative training of models across multiple institutions, particularly for medical image classification tasks.

The paper proposes a federated learning architecture that combines edge computing and 6G networks, utilizing SMPC protocols to encrypt local models and perform encrypted aggregation in the cloud. Additionally, it explores methods for encrypted inference to ensure data security in the cloud environment. The effectiveness of this framework is validated through experiments, and future research directions are discussed, such as the construction of 6G infrastructure, lightweight machine learning models, anomaly detection mechanisms, and communication cost optimization.