Designing Robust LMCA-based Threshold Secret Sharing Scheme for Digital Images Using Multiple Configurations Assignment

Abstract

In this paper, we present a new (t,n)-threshold secret images sharing scheme based on linear memory cellular automata (LMCA). While all existing LMCA-based sharing scheme are not robust, the proposed one provides full robustness property. Precisely, any subset of t participants can collude to recover the shared secret, in contrast to existing LMCA-based schemes when this is possible only for participants having consecutive shares. To achieve robustness, produced shares are constructed using subsets of different LMCA’s configurations instead of using single ones. The subsets are defined according to an assignments matrix that is generated using a specific heuristic. The proposed scheme is shown to be robust, and its security is experimentally evaluated with respect to the problem of secret color image sharing. Obtained results illustrate the secrecy of the produced shares, while comparison gives an accurate evaluation with respect to existing schemes.