User Rank: A User Influence-Based Data Distribution Optimization Method for Privacy Protection in Cloud Storage System

Chang Guo, Qingni Shen, Yahui Yang, Zhonghai Wu

Abstract: The spring up of cloud storage, such as Hadoop HDFS, Open Stack Swift, brings us more intelligent storage solutions. Nowadays, the most commercial version of cloud storage system puts more emphasis on high-performance and high-availability, very little attention is given to privacy protection. This paper proposes a user influence-based data distribution optimization method, User Rank, which migrate blocks of high-impact data from threatened nodes to secure nodes. Additionally, a reference implementation of User Rank is presented to optimize cloud storage system. The simulation experiments results verify that User Rank can significantly reduce private leakage for private or public cloud.