Affiliations: School of Management Science and Engineering, Shandong University of Finance and Economics, Jinan, China
Correspondence:
[*]
Corresponding author. Wu Xiuguo, School of Management Science and Engineering, Shandong University of Finance and Economics, 250014 Jinan, China. Tel.: +86 13064002238; Fax: +86 5318852229; E-mail: xiuguosd@163.com.
Abstract: Replication technology is one of the most effective measure to improve the cloud data availability and reduce the response time, by put multiple copies on different data centers in the cloud. However, traditional replica placement strategies mainly focused on the performance improvement, such as transfer time reduction, load balancing and so on, with little attention to their safety storage, which inevitably result in new security risks towards replicas’ correct access. Aimed to develop an intelligent solution to enhance the cloud storage system performance, we apply fuzzy comprehensive evaluation to candidate data centers, evaluating its security risk level as replica placements, therefore improve the quality of cloud storage system. First, a novel five-dimensional security model for replica placement is proposed. Then, a security-aware replica placement evaluation algorithm is designed and implemented based on fuzzy set and entropy weight theory to establish a fusion of multiple factors of assessment model. Through comparing the experiment results with the outcome of the different strategies, we can see that data replica availability probability of the proposed strategy is up to 85% and less wait latency can be achieved, indicating the better effectiveness and practicality.
