Open Access

ARTICLE

Privacy Data Management Mechanism Based on Blockchain and Federated Learning

Mingsen Mo¹, Shan Ji², Xiaowan Wang^3,*, Ghulam Mohiuddin⁴, Yongjun Ren¹
1 Engineering Research Center of Digital Forensics of Ministry of Education, School of Computer, Nanjing University of Information Science & Technology, Nanjing, 210044, China
2 Zhengde Polytechnic, Nanjing, 211106, China
3 Xi’an University of Posts & Telecommunications, Xi’an, 710061, China
4 Department of Cyber Security at VaporVM, Abu Dhabi, 999041, United Arab Emirates
* Corresponding Author: Xiaowan Wang. Email:

Computers, Materials & Continua 2023, 74(1), 37-53. https://doi.org/10.32604/cmc.2023.028843

Received 19 February 2022; Accepted 19 April 2022; Issue published 22 September 2022

Abstract

Due to the extensive use of various intelligent terminals and the popularity of network social tools, a large amount of data in the field of medical emerged. How to manage these massive data safely and reliably has become an important challenge for the medical network community. This paper proposes a data management framework of medical network community based on Consortium Blockchain (CB) and Federated learning (FL), which realizes the data security sharing between medical institutions and research institutions. Under this framework, the data security sharing mechanism of medical network community based on smart contract and the data privacy protection mechanism based on FL and alliance chain are designed to ensure the security of data and the privacy of important data in medical network community, respectively. An intelligent contract system based on Keyed-Homomorphic Public Key (KH-PKE) Encryption scheme is designed, so that medical data can be saved in the CB in the form of ciphertext, and the automatic sharing of data is realized. Zero knowledge mechanism is used to ensure the correctness of shared data. Moreover, the zero-knowledge mechanism introduces the dynamic group signature mechanism of chosen ciphertext attack (CCA) anonymity, which makes the scheme more efficient in computing and communication cost. In the end of this paper, the performance of the scheme is analyzed from both asymptotic and practical aspects. Through experimental comparative analysis, the scheme proposed in this paper is more effective and feasible.

---

Keywords

Data management; blockchain; federated learning

---