On Demand Ciphering Engine Using Artificial Neural Network

Qutaiba Ali and Shefa Dawwd

Department of Computer Engineering, Mosul University, Iraq

Abstract: In this paper, a new light weight highly secured ciphering engine creation methodology we called On Demand Ciphering Engine (ODCE) was suggested. The main feature of this method is that both, the ciphering engine and the secret key, are kept secrets and created by the user or the system administrator when initiating his transmission, then the involved sides exchange these secrets. The design methodology and structure of this system was described and prototyped using Artificial Neural Network (ANN) as the main building block in the system. Software and hardware implementations of the suggested system were performed to prove its practicability. Also, different experimental tests were achieved to determine the impact of the suggested method on both network delay and system performance.

Keywords: Network security, cryptography, neural network parallel implementation, ODCE, normalized ciphering engine.

Received November 22, 2013; accepted October 26, 2014