Dhara Chirag, de la Torre Gonzalo, Acín Antonio
ICFO-Institut de Ciencies Fotoniques, E-08860 Castelldefels, Barcelona, Spain.
ICFO-Institut de Ciencies Fotoniques, E-08860 Castelldefels, Barcelona, Spain and ICREA-Institucio Catalana de Recerca i Estudis Avançats, E-08010 Barcelona, Spain.
Phys Rev Lett. 2014 Mar 14;112(10):100402. doi: 10.1103/PhysRevLett.112.100402. Epub 2014 Mar 10.
In general, any observed random process includes two qualitatively different forms of randomness: apparent randomness, which results both from ignorance or lack of control of degrees of freedom in the system, and intrinsic randomness, which is not ascribable to any such cause. While classical systems only possess the first kind of randomness, quantum systems may exhibit some intrinsic randomness. In this Letter, we provide quantum processes in which all the observed randomness is fully intrinsic. These results are derived under minimal assumptions: the validity of the no-signaling principle and an arbitrary (but not absolute) lack of freedom of choice. Our results prove that quantum predictions cannot be completed already in simple finite scenarios, for instance of three parties performing two dichotomic measurements. Moreover, the observed randomness tends to a perfect random bit when increasing the number of parties, thus, defining an explicit process attaining full randomness amplification.
一般来说,任何观察到的随机过程都包括两种性质不同的随机性形式:表观随机性,它源于对系统自由度的无知或缺乏控制;以及内在随机性,它不能归因于任何此类原因。虽然经典系统只具有第一种随机性,但量子系统可能表现出一些内在随机性。在本信函中,我们提供了所有观察到的随机性都是完全内在的量子过程。这些结果是在最小假设下推导出来的:无信号传递原理的有效性以及任意(但非绝对)的选择自由缺失。我们的结果证明,量子预测在简单的有限场景中就已经无法完成,例如三方进行两次二分测量的场景。此外,当增加参与方数量时,观察到的随机性趋向于完美的随机比特,因此,定义了一个实现完全随机放大的明确过程。