Secured data partitioning through sequence based mapping and random order of data separation

Abstract

Data partitioning using secret sharing is a popular technique for securing data outsourcing in cloud computing. However, its complexity in reconstructing while preserving confidentiality has limited it for practical use. The drawback of secret sharing on how effectively reconstruct the secret shares, especially when it involves big data has motivated us to propose a sequence based mapping. Furthermore, in the current practice, the generated shares are being sent and stored in the original order which they are being generated. Thus, this could expose them to various threats, if attackers or curious server learn and observe the orders. Therefore, we have presented random order data separation to generate the random order of generating shares. This technique allows data to be separated into multiple chunks, and distributed to cloud storage in random orders. For evaluation, the proposed techniques have been evaluated through a series of simulation using maximum 10000 data. The performance was evaluated based on the time taken to achieve data reconstruction. As a result, we proved that sequence based mapping technique has improved the performance of data reconstruction compared to the indexing technique. In conclusion, sequence based mapping and random order of separated data are the ideal combinations for improving performance and preserving the confidentiality of data in cloud computing.