The privacy and confidentiality of our data are more fundamental to our existence than ever, especially with our growing dependence on online services and the transfer of information, be it for convenience or the unfortunate COVID 19 pandemic. Traditional security measures such as cryptography and steganography are essential. However, they are not sufficient anymore as when one party maintains complete control over the data, they may intentionally, or not, lose or reveal it. Thus, secret sharing schemes were introduced to address applications that require collective agreement by authorized personnel to access or modify, such as missile launching, sophisticated medical agreements, or high-level bank transfers. In standard secret sharing, the target key is distributed among several authorized participants in a way that only the intended group of them are needed to reconstruct the original target key. These schemes became popular as they fortified the security of both cryptography and steganography and accomplished extraordinary results combined with each. This work focuses on two particular secret sharing techniques known as counting-based secret sharing and matrix-based secret sharing, which is based on the former. These methods are simple and intuitive. Consequently, they are prone to attacks that may result in the successful guessing of the key. In this work, not only are the shares hidden but they are also encrypted beforehand so that should they be intercepted, the adversary cannot decipher them. In other words, two layers of security are added to the secret-sharing method: steganography and cryptography. We studied two image steganography methods: least significant bit (LSB) and discrete wavelet transform (DWT), each combined with XOR encryption for security and robustness acceptability verification. The research results demonstrated that the use of steganography and encryption along with the matrix-based secret sharing did not affect the quality of operation nor compromised the security of information presenting attractive remarks.
Read full abstract