This thesis explores quantum random number generators (QRNG) as advanced sources of randomness that overcome the limitations of classical approaches. First, the theoretical background and the role of randomness in cryptography, simulations, and scientific research are presented. Special attention is given to the differences between pseudorandom number generators (PRNG), true random number generators (TRNG), and quantum random number generators (QRNG), with the latter relying on inherently nondeterministic quantum phenomena.
In the experimental part, I analyze the operation of the Qocka device based on single-photon splitting and discuss the acquisition of raw data. The data were processed using von Neumann and Toeplitz extractors and evaluated with statistical tests (Dieharder, PractRand) as well as min-entropy estimation (NIST SP 800-90B). The results show that processed sequences pass stringent tests, but this alone does not guarantee genuine entropy, since a deterministic generator can achieve similar statistical properties. I conclude that reliable evaluation of QRNG requires a combination of statistical testing, source modeling, and conservative entropy estimation. The thesis thus contributes to understanding the importance of post-processing and assessment methods in the development of secure quantum generators.
|
