Adaptive Feedback Compensation Algorithm for Quantum Random Number Generators

As a core component in quantum cryptography, Quantum Random Number Generators (QRNGs) face dual critical challenges: insufficient randomness enhancement and limited compatibility with post-processing algorithms. This study proposes an Adaptive Feedback Compensation Algorithm (AFCA) to address these limitations through dynamic parameter feedback and selective encryption strategies. The AFCA dynamically adjusts nonlinear transformation intensity based on real-time statistical deviations, retaining over 50% of original bits while correcting local imbalances. Experimental results demonstrate significant improvements across QRNG types: the Monobit Test <i>p</i>-value for continuous QRNGs increased from 0.1376 to 0.9743, and the 0/1 distribution deviation in discrete QRNGs decreased from 7.9% to 0.5%. Compared to traditional methods like von Neumann correction, AFCA reduces data discard rates by over 55% without compromising processing efficiency. These advancements provide a robust solution for high-security quantum communication systems requiring multi-layered encryption architectures.