Secure Random Number Generation in Continuous Variable Systems

Abstract

Intrinsic uncertainty is a distinctive feature of quantum physics, which can be used to harness high-quality randomness. However, in realistic scenarios, the raw output of a quantum random-number generator (QRNG) is inevitably tainted by classical technical noise. The integrity of such a device can be compromised if this noise is tampered with, or even controlled by some malicious parties. In this chapter, we first briefly discuss how the quantum randomness can be characterised via information theoretic approaches, namely by quantifying the Shannon entropy and min-entropy. We then consider several ways where classical side-information can be taken into account via these quantities in a continuous-variable QRNG. Next, we focus on side-information independent randomness that is quantified by min-entropy conditioned on the classical noise. To this end, we present a method for maximizing the conditional min-entropy from a given quantum-to-classical-noise ratio. We demonstrate our approach on a vacuum state CV-QRNG. Lastly, we highlight several recent developments in the quest of developing secure CV-QRNG.