Cascade Forward Back-propagation Neural Network Based Group Authentication Using (n, n) Secret Sharing Scheme

Abstract

Authentication is the important issue over the internet. It results in need of robust security services and schemes. This paper proposes the Shamir Secret Sharing Scheme along with Cascade forward back propagation neural network. The proposed (n, n) Shamir secret sharing scheme is implemented successfully for share encryption and decryption process. It reveals the secret only and only when all the n number of participants are available at the reconstruction process. The generated shares have a best quality so that human brain can’t predict the original secret by any combination. The neural network is already trained for RGB image database and images are stored in particular group having a group category number which is to be predicted as output of neural network. For testing purpose, we take an image, store it in a group, create its shares and now identifies to which group it belongs. The neural network here provides the correct group of that image which we have trained earlier.