Public verification of data integrity is crucial in promoting the serviceability of cloud storage systems. Auditing technique is an efficient tool to check the integrity of the remotely stored data. However, the general audit mechanism only considers the confidentiality of outsourced data, and ignores the anonymity of data owners. The exposure of data ownership causes outsourced data become vulnerable to attacks. To address this problem the study presents a new model of pseudo-ID-based auditing protocol. We propose the first pseudo-ID-based public auditing scheme (PIDPA) for cloud data integrity, which is provably secure based on the computational Diffie-Hellman (CDH) problem. In particular, the introduction of pseudo-identity ensures the anonymity of data owners and allows a key generation center (KGC) to track the real identity of data owners when disputes arise. Moreover, full batch auditing is supported in the multi-user scenario on the basis of homomorphic signature. Theoretical analyses prove that the proposed PIDPA scheme is secure while maintaining the desirable security properties, and experimental results show that our scheme is efficient to audit data integrity in the cloud.