Research on Data Security Isolation and Sharing Mechanism in Cloud Multi-tenant Environment Based on Blockchain and Federated Learning

Guangwei Li; Wenjing Fang; Kaifu Mao

doi:10.54097/4ypqxn17

Authors

Guangwei Li
Wenjing Fang
Kaifu Mao

DOI:

https://doi.org/10.54097/4ypqxn17

Keywords:

Cloud Multi-Tenancy, Data Security Isolation, Federated Learning, Blockchain, Adaptive Access Control.

Abstract

In cloud computing multi-tenant environments, the contradiction between data security isolation and efficient sharing has become increasingly prominent. This paper proposes an innovative framework integrating blockchain technology and federated learning to address the imbalance between security isolation strength and data circulation efficiency in traditional technologies. Through dynamic encryption strategies, adaptive access control algorithms, and cross-tenant joint modeling methods, an integrated design of "secure isolation-intelligent sharing-trustworthy auditing" is achieved. A multi-dimensional permission management model based on attribute graphs is proposed, which dynamically optimizes access policies using machine learning. The tamper-proof feature of blockchain is leveraged to record the full lifecycle of data operations, enhancing the transparency of sharing processes. Federated learning technology is employed to enable secure joint analysis with the principle of "data in place, models moving." Experiments in a simulated medical cloud environment show that the proposed mechanism reduces data leakage risks by 40%, improves sharing efficiency by 30%, and increases the accuracy of federated learning models by 15% compared with traditional solutions, providing a solution with both theoretical innovation and engineering value for data security governance in multi-tenant environments.

References

[1] Armando A, Bifulco R, Castiglione A, Santis A D. A survey of isolation techniques in cloud computing[J]. IEEE Communications Surveys & Tutorials, 2020, 22(3): 1863–1885.

[2] Zhang Y, Li H, Wang X, Chen J. Logical isolation technology for multi-tenant cloud storage[J]. IEEE Transactions on Parallel and Distributed Systems, 2021, 32(5): 1215–1228.

[3] Li X, Wang Q, Li Y, Zheng Z. A hybrid isolation framework for multi-tenant cloud environments[J]. Journal of Cloud Computing, 2022, 11(1): 1–20.

[4] Zheng Z, Xie S, Dai H, Chen X, Wang H. Challenges and opportunities of blockchain: A survey[J]. IEEE Transactions on Industrial Informatics, 2018, 14(4): 1624–1634.

[5] European Parliament & Council of the European Union. Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data (General Data Protection Regulation)[S]. Official Journal of the European Union, 2016, L 119: 1–88.

[6] Standing Committee of the National People’s Congress. Data Security Law of the People’s Republic of China[S]. 2021.

[7] Burns B. Kubernetes in action[M]. Manning Publications, 2020.

[8] Tsoutsos G, Asimakopoulos E, Ioannidis S. A survey of container escape vulnerabilities[J]. IEEE Access, 2021, 9: 11489–11506.

[9] McMahan B, Moore E, Ramage D, Hampson S, y Arcas B A. Federated learning of deep networks from decentralized data[C]. Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS), 2017: 1273–1282.

[10] Kairouz P, McMahan B, Avent H, et al. Advances and open problems in federated learning[J]. Foundations and Trends® in Machine Learning, 2019, 12(1): 1–149.

[11] Geyer A, Kleinberg J, Nabi M. Differentially private federated learning: Client-level differential privacy[C]. Proceedings of the 2017 ACM SIGSAC Conference on Computer and Communications Security, 2017: 1156–1171.

[12] Ma Y, Zhang Q, Zheng Z, Chen X. Privacy-preserving and verifiable federated learning via secure multi-party computation[J]. IEEE Transactions on Industrial Informatics, 2022, 18(9): 6066–6075.

[13] Martinez A B. Blockchain-based incentives for federated learning[R]. arXiv preprint arXiv:2104.03231, 2021.

[14] Majeed A, Hong X. FLchain: Federated learning meets blockchain[J]. Future Generation Computer Systems, 2022, 131: 37–47.

[15] Intel. Intel SGX: Enabling trusted cloud computing[R]. 2025.

[16] Liu X, Lyu M R, Jin H. Encrypted model aggregation for privacy-preserving federated learning[J]. IEEE Transactions on Information Forensics and Security, 2021, 16: 356–369.

[17] Li X, Wang Y, Zhang H. A secure medical data sharing scheme based on federated learning with homomorphic encryption[J]. Mathematical Biosciences and Engineering, 2023, 20(2): 1543–1562.

[18] Bank for International Settlements. Homomorphic encryption in financial data sharing: Challenges and applications[J]. BIS Quarterly Review, 2024, (3): 45–58.