You'd have to break the laws of physics to break this code
Scientists are a step closer to creating secret codes that are absolutely unbreakable, advancing hopes for protecting sensitive data from any kind of computer attack. Three independent research groups* are simultaneously reporting, in the American Physical Society's peer-reviewed journal Physical Review Letters (tentatively scheduled for publication in May), the first demonstrations of sending encrypted messages using "quantum entanglement."
Quantum entanglement, which Einstein called "spooky action at a distance," refers to particles that -- even when very far apart -- are intimately linked. Measurement of one entangled light particle (a photon), then, determines the properties of the other. Since quantum mechanics says that -- until measured -- a particle's basic properties can be in a combination of states, a code derived from entangled photons could stay truly secret until "read" by both a sender and receiver. Furthermore, eavesdropping attempts could be easily detected, since they would unavoidably disturb the photon's properties. In effect, the code is protected by the laws of nature.
Standard quantum cryptography, in which the sender creates and sends code made from a series of individual photons -- with different polarizations (the direction in which a photon's electric field vibrates) representing the 0's and 1's of computer language -- has been demonstrated before. But there have been drawbacks. Most of the time the very faint pulses used contain no photons at all, which is highly inefficient. In addition, sometimes the pulses contain more than one photon, in which case a technologically advanced eavesdropper could skim photons from the signal and covertly gain information about the secret key.
Quantum cryptography using entangled pairs of photons allows easier detection if a photon is "stolen." Also, the entanglement process generates an inherently random code while allowing the use of brighter pulses. The net result could eventually be a higher transmission rate, over longer distances, with greater security.
The data encryption currently used, in electronic banking for example, relies on a secret code that utilizes a number with more than 100 digits. To break the code, one needs to factor that large number into two smaller numbers. This is an extremely demanding task, and currently can't be done in a reasonable amount of time by even the world's most state-of-the-art supercomputers. However, a mathematical breakthrough or development of an advanced processor (such as quantum computers which could process many mathematical calculations simultaneously) may eventually be able to crack such a code -- instantly causing a threat to privacy and national security by making computer systems insecure.
"Admittedly, that day is probably decades away," says Artur Ekert, a physicist at the University of Oxford in England who conceived of the use of entanglement in quantum cryptography. "But can anyone prove, or give any reliable assurance, that it is? Confidence in the slowness of technological progress is all that the security of the (current) system now rests upon."
In the most basic version of entangled quantum cryptography, a specially prepared crystal splits a single photon into a pair of entangled photons. When this occurs, according to quantum mechanics, the polarization of each photon in undetermined -- simultaneously representing a mixture of both 0 and 1. Only when one of the photons is measured, or otherwise disturbed, does it gain a definite polarization and represent a particular value. Whether it becomes a 0 or 1 is completely random, but once its polarization is determined, the other photon is forced to assume a polarization that's correlated with its entangled partner. When using the same detector setting on each end (selected randomly, as well, from a couple of options), sender and receiver get the same polarization -- and digit. After receiving a series of entangled photons, sender and receiver can discuss the settings used -- rather than the actual readings -- over a public channel, like a telephone line or the Internet.
After discarding readings made when the sender and receiver weren't using the same settings, they have a randomly generated string of digits that can serve as a completely secure secret key for encoding, and de-coding, the actual message.
According to David DiVincenzo of the IBM Watson Research Center, the technical ability for widespread practical application of entangled cryptography is a long way off. Nevertheless, he says, "this is a small, but notable, step towards qualitatively more powerful forms of long-distance cryptography."
*For copies of the Physical Review Letters papers, contact: Ben Stein American Institute of Physics 301-209-3091 bstein@aip.org
Randy Atkins American Physical Society 301-209-3238 atkins@aps.org
Author contacts for each of the three papers: Paul Kwiat Los Alamos National Laboratory 505-667-6173 Kwiat@lanl.gov
Thomas Jennewein University of Vienna 43-4277-51207 (note: this involved collaboration with the University of Munich)
Nicolas Gisin University of Geneva 41-22-702-65-97 Nicolas.Gisin@physics.unige.ch
Independent sources used in the above release: Artur Ekert University of Oxford a.ekert1@physics.ox.ac.uk
David DiVincenzo IBM Watson Research Center 914-945-4421 divince@watson.ibm.com
Journal
Physical Review Letters