Desenvolvimento e avaliação de um gerador de números verdadeiramente aleatórios baseado em ruído acústico

Abstract

This work presents the development and evaluation of a low-cost true random number generator (TRNG) based on the capture of environmental acoustic noise using a MAX9814 microphone and an ESP32 microcontroller. The methodology focused on extracting bits from analog-to-digital converter (ADC) samples and applying the Peres Extractor as postprocessing, while investigating the influence of environmental variables monitored by the BME280 sensor. The results showed that, although the raw bit sequences exhibited statistical bias and only partial approval in the tests, post-processing eliminated the dependence on climatic conditions and raised the min-entropy to values above 0.997, ensuring approval in all 15 tests of the NIST SP 800-22 suite. The solution proved to be a viable, robust, and low-cost alternative for generating truly random numbers.