Three-level storage and nested MDS codes for perfect secrecy in multiple clouds
Ping Hu, Chi Wan Sung, Siu‐Wai Ho, Terence Chan · 2014
The problem of storing data reliably and securely in multiple cloud storage providers (CSPs) with minimum cost is investigated. A jointly optimal coding and storage allocation scheme, which achieves perfect secrecy with minimum cost, is derived. The optimal coding scheme is shown to be the nested maximum-distance-separable code and the optimal amounts of data to be stored in the CSPs is proven to exhibit a three-level structure. The exact parameters of the code and the exact storage amount to each CSP can be determined numerically by simple one-dimensional search.