A Quantum Key Distribution protocols emulated post-processing method

Abstract Random numbers generated by unpredictable entropy sources and deterministic algorithms are widely used in important applications. However, raw random numbers frequently exhibit statistical defects. To address the issue, post-processing methods are employed to refine raw random numbers and ensure they are compliant with statistical requirements. In this study, we propose a novel post-processing method that emulates quantum key distribution protocols to theoretically eliminate the bias without length loss. Specifically, the polarizer, filter, and decider are constructed using uniformly distributed random seeds. The polarizer and filter implement confusion and asymmetric transformations, and the decider performs substitution operations. Simultaneously, it leverages iterative processing to progressively improve the statistical quality of raw random numbers. Experimental results illustrate the effectiveness and efficiency of proposed method by means of statistical analyses (entropy, autocorrelation, and NIST SP 800-22 randomness test, etc.) along with a processing time at 3 seconds for 2.4 × 10⁸ bits.