Analyses of unpredictable properties of a wind-driven triboelectric random number generator

Abstract

Wind-driven triboelectric nanogenerators (W-TENGs) are a promising candidate for an energy harvester because wind itself possesses unexhausted, ubiquitous, and clean properties. W-TENG has also been used as a random number generator (RNG) due to the inherent chaotic properties of wind that is also an entropy source. Thus, a W-TENG which simultaneously generates both power and true random numbers with a two-in-one structure, is a wind-driven RNG (W-RNG) like the Janus. However, a root cause of W-RNG unpredictability has not been elucidated. In this work, the unpredictability, which is essential and critical for an RNG, is statistically and mathematically analyzed by auto-correlation, cross-correlation, joint entropy, and mutual information. Even though the overall shape of the total output analog signals from the W-RNG looks like a sinusoidal wave that is not obviously unpredictable, discretized digital signals from the continuous analog output become unpredictable. Furthermore, partial adoption of 4-bit data from 8-bit raw data, with the aid of analog-to-digital converter hardware, further boosts the unpredictability. The W-RNG, which functions as a W-TENG, can contribute to self-powering and self-securing outdoor electrical systems, such as drones, by harvesting energy and generating true random numbers.