Study on post-processing algorithms for true random number generators

Randomness is essential for cryptographic applications. The true random number generators (TRNGs) produce true randomness based on unpredictable physical phenomena and offer a stronger security guarantee than pseudorandom number generators (PRNGs). However, hardware entropy sources rarely provide perfect random numbers, postprocessing is required for TRNGs to overcome potential statistical defects. In this paper, we introduce and analyze four conventional lightweight post-processing algorithms. Then, we perform an in-depth study on the hash function based postprocessing from the applicability perspective.