A Post-Processing-Free True Random Number Generator Based on a Chaotic Piezoelectric MEMS Oscillator
Journal
IEEE Electron Device Letters
ISSN
07413106
Date Issued
2025
Author(s)
Abstract
Hardware true random number generators (TRNGs) based on chaotic MEMS oscillators often face practical limitations, including operational hysteresis, high driving voltages, and a heavy reliance on statistical post-processing to ensure randomness. This work overcomes these challenges by demonstrating a TRNG that uses amplitude modulation (AM) to induce chaotic internal resonance (IR) in a ScAlN piezoelectric oscillator. The AM-driven dynamics transform the oscillator's periodic beating into a robust, aperiodic waveform, providing a direct source of high-quality physical entropy. Without any statistical post-processing, the resulting 1.512-Mbit bitstream at a bit rate of 540-kbps passes all NIST 800-22 and 800-90B independent and identically distributed (IID) tests. The bitstream also shows a non-IID min-entropy of 0.843885 and an autocorrelation function (ACF) value of ±0.00159 within the 95% confidence interval (CI), confirming the AM-driven IR oscillator as a simple, area-efficient, and robust pathway for on-chip cryptographic hardware.
Subjects
amplitude modulation
chaotic internal resonance
Nonlinear oscillator
true random number generator
Publisher
Institute of Electrical and Electronics Engineers Inc.
Type
journal article