Characterizing Environmental Thermal Fluctuations for Use in True Random Number Generators

Abstract

Random number generation is significant for secure communication, data security and cryptography. However, while Quantum Random Number Generators (QRNGs) rely on quantum uncertainty, environmental data can also serve as practical sources of randomness for true random number generation. Environmental data such as temperature, humidity, and wind speed exhibit continuous variability over time; these changes arise from complex weather behavior. In this paper, temperature data collected from two meteorological stations in Pakistan, Karachi and Hyderabad, are used to generate random bits. Everyday temperature values are transformed into binary sequences using a mean-based thresholding technique, followed by post-processing with the Von Neumann extractor to decrease bias and correlation. The quality of the generated random bits is evaluated using Shannon entropy, lossless compression testing with the Gzip, Bzip2, and LZMA algorithms, the NIST SP 800-22 statistical test suite and the auto-correlation analysis. The results determine that correctly processed temperature-based entropy can produce statistically usable random sequences suitable for randomness testing and security-related research.