A Novel Randomness Source for Cryptographic Applications Based on Perceptron-Driven Chaotic Maps

Abstract

The generation of robust binary random sequences is one of the most widely used cryptographic primitives in the field of cryptography and data security in general. It constitutes the fundamental basis of most cryptographic and security protocols. Consequently, the robustness and efficiency of these systems strongly depend on the quality of random number generators, which must produce unpredictable and statistically robust sequences in order to protect data transmitted over insecure channels such as the Internet, Internet of Things (IoT) networks, or stored in remote environments such as the cloud. Several techniques have been proposed for the design of random number generators, including cellular automata, linear feedback shift registers (LFSRs), and chaotic maps. The objective of this work is to design a new robust pseudo-random number generation method that can be used as a building block for cryptographic primitives and protocols. The proposed method is based on a chaotic map, which is characterized by strong sensitivity to initial conditions. The main idea is to extend this dynamical system in order to overcome its limitations, such as its narrow chaotic interval, by integrating it with a perceptron model. This hybrid approach leads to a more intelligent system with a significantly larger key space. To prove the robustness and applicability of the designed generator in cryptographic contexts, several statistical analyses and tests were conducted, including entropy analysis, histogram evaluation, avalanche effect measurement, and the NIST statistical test suite, which is considered a standard in the field of cryptography. The outcoming results are powerfully satisfactory and demonstrate the robustness of the generated sequences, confirming their suitability as a reliable source of randomness for secure cryptographic protocols.