---

GIT-CERCS-05-04
Arun Subbiah, Douglas Blough,
Practical Share Renewal for Large Amounts of Data

Threshold secret sharing schemes encode data into several shares such that a threshold number of shares can be used to recover the data. Such schemes provide confidentiality of stored data without using encryption, thus avoiding the problems associated with key management. To provide long-term confidentiality, proactive secret sharing techniques can be used, where shares are refreshed or renewed periodically so that an adversary who obtains fewer than the threshold shares in each time period does not learn any information on the encoded data. Share renewal is an expensive process, in terms of the computation and network communication involved. In the proactive model, this share renewal process must complete as soon as possible so that an adversary who compromises servers in the present time period does not learn shares stored in the last time period.

This paper proposes an algorithm where the shares of all the stored data are renewed by the share renewal of only one secret. The computation and network communication overheads are thus drastically reduced, allowing for the share renewal of all the stored data to complete quickly. These benefits are gained at the expense of some performance penalty during reads and writes, which is shown to be worthwhile.

 

---