Fuzzy Identity-Based Data Integrity Auditing for Reliable Cloud Storage Systems

TLDR

Data integrity is a critical security concern in cloud storage, and auditing protocols enable verifiers to check outsourced data without downloading it, yet existing designs suffer from complex key management. This work introduces the first fuzzy identity‑based auditing scheme to simplify key management in cloud data integrity checking. The authors define a primitive where a user’s identity is a set of descriptive attributes, formalize its system and security models, and construct a concrete protocol that uses biometric attributes as fuzzy identities. The protocol achieves error‑tolerance by allowing verification between sufficiently similar identities, is provably secure under CDH and DL assumptions, and a prototype confirms its practical feasibility.

Abstract

Data integrity, a core security issue in reliable cloud storage, has received much attention. Data auditing protocols enable a verifier to efficiently check the integrity of the outsourced data without downloading the data. A key research challenge associated with existing designs of data auditing protocols is the complexity in key management. In this paper, we seek to address the complex key management challenge in cloud data integrity checking by introducing fuzzy identity-based auditing, the first in such an approach, to the best of our knowledge. More specifically, we present the primitive of fuzzy identity-based data auditing, where a user's identity can be viewed as a set of descriptive attributes. We formalize the system model and the security model for this new primitive. We then present a concrete construction of fuzzy identity-based auditing protocol by utilizing biometrics as the fuzzy identity. The new protocol offers the property of error-tolerance, namely, it binds with private key to one identity which can be used to verify the correctness of a response generated with another identity, if and only if both identities are sufficiently close. We prove the security of our protocol based on the computational Diffie-Hellman assumption and the discrete logarithm assumption in the selective-ID security model. Finally, we develop a prototype implementation of the protocol which demonstrates the practicality of the proposal.