A Performance Metric for Discrete-Time Chaos-Based Truly Random Number Generators

TL;DR

This paper introduces an information entropy-based metric to evaluate and improve the statistical quality of binary sequences generated by discrete-time chaos-based TRNGs, aiding design and post-processing.

Contribution

It proposes a novel entropy-based metric specifically for assessing and optimizing the performance of chaos-based TRNGs.

Findings

The metric effectively quantifies the statistical quality of TRNG outputs.

It assists in designing better TRNG circuits and post-processing units.

The approach addresses degradation due to process variations.

Abstract

In this paper, we develop an information entropy based metric that represents the statistical quality of the generated binary sequence in Truly Random Number Generators (TRNG). The metric can be used for the design and optimization of the TRNG circuits as well as the development of efficient post-processing units for recovering the degraded statistical characteristics of the signal due to process variations.