Fidelius: A Novel Secure Data Analysis Framework Leveraging Intel SGX and Blockchain

Fidelius: A Novel Secure Data Analysis Framework Leveraging Intel SGX and Blockchain

연구 분야: Cryptography

학회: BSCI '25: Proceedings of the 7th ACM International Symposium on Blockchain and Secure Critical Infrastructure

As the reliance on data analysis grows in the era of big data, concerns over data leakage and privacy breaches have become increasingly prevalent. While existing technologies such as Secure Multi-Party Computation (MPC), Homomorphic Encryption (HE), Federated Learning (FL), and Trusted Execution Environments (TEE) aim to address these concerns, they often exhibit limitations in addressing complex data analysis scenarios involving multiple roles. In this paper, we propose Fidelius, a novel system that leverages Intel SGX and blockchain to enhance data analysis security. Fidelius employs a static binary analysis approach and a privacy description language (PDL) to prevent data leakage in computation results. Furthermore, it introduces a cryptographic protocol to ensure the trustworthiness and verifiability of computation results, along with a combination of cryptographic protocol and local attestation to achieve consistent verification of analysis programs. Experimental results demonstrate that Fidelius incurs minimal overhead while surpassing existing solutions in performance. Thus, Fidelius presents a promising solution to enhance the security of data analysis in complex scenarios involving multiple roles.