A novel group based cryptosystem based on electromagnetic rotor machine

Abstract: In this paper, an algorithm is aimed to make a cryptosystem for gray level images based on voice features, secret sharing scheme and electromagnetic rotor machine. Here, Shamir secret sharing (k n) threshold scheme is used to secure a key along with voice features of (n k) users. Keystream is molded by coefficients of a voice sample, using this key stream, rotor machines rotating cylinders positions are initialized and internal wiring is decided by pseudo random number of Henon chaotic map, where initial seed for chaotic system is chosen from keystream. And furthermore, shares of key stream are distributed among users. Speech processing is fused with electromagnetic machine to provide authentication as well as group based encryption. Perceptual linear predication (PLP) coefficients are utilized for formation of secret key. Simulation experiments and statistical analysis demonstrate that the proposed algorithm is sensitive to initial secret keystream, entropy, mean value analysis and histogram of the encrypted image is admirable. Hence, the proposed scheme is resistible to any vulnerable situation.