Verification of finite size continuous-variable quantum key distribution under coherent attacks

量子密钥分发(QKD)旨在分发秘密信息，在最简单的情况下，只知道给通常称为Alice和Bob的两个直接相关方。与经典密码学不同，量子力学的基本定律确保了以量子态编码的信息不会被复制而不被注意。它被广泛认为是量子信息科学中最成熟的分支，甚至可以进行一些商业实现。该项目的目标是建立一个使用纠缠激光光束的量子密码系统，该系统在几个重要方面改进了早期的工作。我们将第一次建立一个连续变量(CV)QKD协议，提供针对最一般的、一致的攻击的安全性。因此，与以前的CV建议和实现不同的是，新协议不需要强烈的附加假设来约束攻击类型，而是通过一个渐近公式来评估量子错误率。这个项目的理论合作伙伴最近才给出了第一个没有这种约束的安全性证明。我们将利用创纪录水平的孪生激光光束之间的纠缠，这只在实验伙伴的实验室中才能获得。这是达到新安全证明所设定的严格标准的先决条件。前述渐近公式仅断言存在适当的数据处理来生成密钥。因此，在以前的大多数实验中，处理步骤往往被视为理所当然。我们将实现它，从而生成实际的密钥。该项目扩展了两个合作伙伴早先的联合工作，在该工作中，计划中的双激光束源的前驱被证明满足渐近安全标准。

Quantum key distribution (QKD) aims at the distribution of secret information, only known, in the simplest case, to two directly involved parties usually called Alice and Bob. In contrast to classical cryptography, fundamental laws of quantum mechanics ensure that the information, encoded in quantum states, cannot be copied unnoticed. It is widely considered the most mature branch of quantum information science, where even some commercial implementations are available. The aim of this project is to build a quantum cryptography system with entangled laser beams, which improves on earlier work in several significant ways. We will for the first time establish a continuous variable (CV) QKD protocol providing security against most general, coherent attacks. Hence, in contrast to previous CV proposals and realizations, which assess security by an asymptotic formula in terms of the quantum error rate, the new protocol does not require strong additional assumptions constraining the type of attack. The first security proof without such constraints was only recently given by the theory partner of this project. We will exploit record levels of entanglement between twin laser beams, which are only available in the lab of the experimental partner. This is a prerequisite for achieving the demanding criteria set by the new security proof. The aforementioned asymptotic formula only asserts the existence of suitable data processing to generate a key. In most previous experiments the processing step was therefore often taken for granted. We will implement it, thus, generate actual key. The project extends earlier joint work of the two partners, in which a precursor of the planned twin laser beam source was shown to meet the asymptotic security criteria.