FACOS: Enabling Privacy Protection Through Fine-Grained Access Control With On-Chain and Off-Chain System

Chao Liu, Cankun Hou, Tianyu Jiang, Jianting Ning, Hui Qiao, Yusen Wu

TL;DR

This paper introduces FACOS, a novel system that combines a permissioned blockchain with off-chain storage to provide fine-grained access control for sensitive data. FACOS enhances privacy and security by integrating attribute-based, broadcast, and threshold encryption methods, ensuring data integrity and availability while mitigating risks of data loss and unauthorized access.

What Problem Does This Paper Address?

In today's data-driven world, especially in sectors like finance, government, and healthcare, the sheer volume of generated information necessitates robust solutions for secure storage, efficient dissemination, and precise access control. While blockchain technology offers decentralized storage and immutability, traditional public blockchains struggle with scalability and privacy for sensitive data. Existing off-chain storage solutions often suffer from single points of failure, leading to data loss and corruption. Furthermore, current access control mechanisms on blockchains can be inefficient, lack fine-grained control, and compromise data privacy. This paper addresses the challenge of building a system that can securely store and manage large volumes of sensitive data, offering granular access control while maintaining privacy and ensuring data availability and integrity.

Methods

FACOS employs a hybrid approach, leveraging both on-chain and off-chain components:

  • Permissioned Blockchain: A consortium blockchain network is used for managing access control policies and transaction metadata, ensuring transparency and immutability.
  • Off-Chain BFT Storage: Asynchronous Byzantine Fault Tolerance (BFT) protocols are utilized for robust and fault-tolerant off-chain data storage, preventing single points of failure and ensuring data availability.
  • Fine-Grained Access Control: FACOS supports three distinct access control methods, allowing data owners to choose the most suitable one based on their needs:
  • Attribute-Based Encryption (ABE): Access is granted based on user attributes.
  • Broadcast Encryption (BE): Efficiently disseminates data to predefined groups.
  • Threshold Encryption (TE): Requires a minimum number of parties to decrypt data, enhancing security.
  • Trusted Execution Environment (TEE): TEEs (specifically using Gramine) are employed to verify client credentials and ensure the integrity of access control policy checks, adding a layer of hardware-based security.
  • Hybrid Encryption: For data confidentiality, FACOS uses symmetric encryption for the message itself and then encrypts the symmetric key using the chosen access control method (ABE, BE, or TE). Hashes of the data are also stored on-chain for integrity verification.

Datasets

Not discussed in the paper. The paper focuses on a system design and evaluation using simulated data and performance metrics.

Key Insights and Ideas

  • Synergy of On-Chain and Off-Chain: The paper demonstrates that a combination of on-chain metadata (hashes, policies) and off-chain robust storage (BFT) is crucial for scalable and secure data management.
  • Flexibility in Access Control: Offering ABE, BE, and TE provides users with a spectrum of choices, allowing them to balance security, efficiency, and group management needs.
  • TEE for Enhanced Trust: Integrating TEEs for credential verification adds a critical layer of trust, especially in scenarios where the verifier needs to be assured of its own integrity.
  • Performance Trade-offs: The evaluation highlights the inherent performance differences between ABE, BE, and TE, showing that BE generally offers the lowest latency due to its reliance on symmetric cryptography, while ABE can be more computationally intensive.
  • Scalability with BFT: The use of asynchronous BFT for off-chain storage proves effective in handling large data volumes and maintaining availability, even in the presence of network failures.

Key Results

  • Efficient Operations: FACOS achieves efficient upload and download times, with reported latencies of approximately 2.79 seconds for uploading and 0.96 seconds for downloading.
  • Comparative Performance: Compared to a referenced scheme [22], FACOS shows significantly better performance, especially for larger data volumes. For 16kb data, FACOS's write and read times were 77.14% and 43.94% of Ref. [22], respectively.
  • Privacy Guarantees: Theoretical analysis proves that FACOS preserves message privacy, hash privacy, and access control policy privacy against PPT adversaries.
  • Robustness: The system is designed to be resilient against Byzantine failures in the off-chain storage and offers protection against DDoS attacks.

Limitations and Future Work

  • TEE Trust Assumption: The system assumes that the TEE and key generation center are fully trusted. Collusion between these entities could pose a security risk.
  • Complexity of ABE: The paper notes that ABE, while powerful for fine-grained control, can be computationally intensive, impacting performance.
  • Future Work: The authors suggest further research into optimizing the performance of ABE and exploring more advanced cryptographic techniques. They also highlight the importance of continued evaluation and real-world deployment to assess the system's long-term viability and scalability.

Summary updated

BibTeX
@article{liu2024facos,
  author = {Chao Liu and Cankun Hou and Tianyu Jiang and Jianting Ning and Hui Qiao and Yusen Wu},
  title = {{FACOS: Enabling Privacy Protection Through Fine-Grained Access Control With On-Chain and Off-Chain System}},
  journal = {IEEE Trans. Inf. Forensics Secur.},
  year = {2024},
  doi = {10.1109/tifs.2024.3427311},
}
Figure