Development of new encryption system using Brownian motion based diffusion

摘要

A new cryptographic model is proposed incorporating intertwining logistic map based confusion process and two dimensional Brownian Motion based diffusion algorithm. An intertwining logistic map is utilized in the algorithm to provide better distribution of random numbers and to overcome blank and stable windows noticed in the schematic of logistic map's bifurcation. Most of the existing image encryptionmodels use raw images without any modification for the confusion and diffusion processes. Their main drawback is that original pixel values remain the same. This problem is overcome by a Pseudo random generator based key stream that modifies pixel values. Further, an intertwining logistic map based confusion process enhances the key sensitivity and complex relationship is created between cipher and test image. Finally, two dimensional Brownian motion based diffusion is applied to bind pixels with each other to such a degree that even one bit modification in the original image affects most of the pixels in the cipher. This makes the model sensitive to change in pixel value or secret key. Various test results indicate that the proposed encryption model can encrypt the plain image into a cipher of random binary sequence. Correlation coefficients among cipher image pixels is found to be negligible, NPCR values are greater than 99.60% (average) and the model is resistive against brute-force, statistical and differential attacks.