Scientists from the National Quantum Information Centre in Gdańsk, in cooperation with a US scientist, have devised a system which is able to extract from seemingly random data the data which is truly and unquestionably coincidental. Their research has been published in the work “Realistic Noise-tolerant Randomness Amplification using Finite Number of Devices” in the prestigious magazine Nature Communications. Such systems for amplifying randomness have been developed before but have proved too difficult to be implemented successfully. The solution put forward by the Poles has much greater chances to be put into use.
In creating random numbers, it is possible to use devices which take data from the outside, from the environment. They are based on, as we assume, stochastic processes (such as, for example, the weather), but we cannot be certain that the processes which at first glance may seem random are in fact so. "In our macroscopic world randomness stems from ignorance. But the fact that we cannot find a rule which governs randomness does not at all mean that such a rule does not exist. Someone may know it. Therefore, there’s never any guarantee that what we are observing is random”, as one of the authors of the research, Prof. Michał Horodecki from the Faculty of Mathematics, Physics and Informatics of the University of Gdańsk told the Polish Press Agency.
The team was the first to show that randomness can be amplified by using a finite number of devices. As stressed by the scientist, 12 devices are sufficient to generate a string of random numbers of any length, from which the fragment of data which should not be trusted has been removed. So far the Polish team has developed only a diagram for such a device and an algorithm for data analysis. The Gdansk scientists are now hoping to progress from theory to experiment and then, who knows, maybe to application.
The research is supported by the Foundation for Polish Science (the TEAM project headed by Prof. Marek Żukowski, the Ministry of Science and Higher Education, and the EU projects RAQUEL and ERC QOLAPS, the latter headed by Prof. Ryszard Horodecki).
The results were obtained in a large team thanks to interdisciplinary cooperation, which is reflected in the team’s make-up of Fernando Brandao (Microsoft); Andrzej Grudka (The Institute of Physics of the Adam Mickiewicz University); Karol Horodecki (The Institute of Informatics of the University of Gdańsk); Michał Horodecki (The Institute of Theoretical Physics and Astrophysics of the University of Gdańsk); Paweł Horodecki (The Faculty of Applied Physics and Mathematics of the Gdańsk Technical University); Ravinshankar Ramanathan (The Institute of Theoretical Physics and Astrophysics of the University of Gdańsk); Tomasz Szarek (The Institute of Mathematics of the University of Gdańsk) and Hanna Wojewódka (The Institute of Theoretical Physics and Astrophysics/ The Institute of Mathematics of the University of Gdańsk).
