Researchers from Stockholm University and Linköping University in Sweden have discovered that quantum cryptography may not be as secure as it was presumed to be. They found that energy-time entanglement, which forms the basis for many systems of quantum cryptography, is vulnerable to attack. The research was published in Science Advances.
Quantum Encryption Entanglements
The energy-time entanglement of quantum encryption is based on testing the connection at the same time as the encryption key is created. In practice, two photons are sent out at exactly the same time in two different directions. At both ends of the connection, an interferometer is placed to measure the interference of the detected light.
If the photon stream is being eavesdropped, there should be observational noise present, which can be detected using a theorem from quantum mechanics called Bell’s inequality. If the connection is secure and free from noise, the remaining photons can be used as an encryption key to protect the message.
Now, one interpretation of quantum mechanics (QM) says two quantities cannot be measured at the same time because the measurement process disturbs the system; think Heisenberg uncertainty, but for QM. This interpretation allowed hidden variables to determine the outcome of an experiment.
Bell’s inequalities provide a way to test this interpretation. If there were hidden variables, then the observed distributions would have to have come from a single, hidden joint distribution and would therefore have to obey Bell’s inequalities.
But as the paper published in Science Advances showed, a specially crafted light source can fool the interferometers, making the Bell’s inequality test fail.
How to Hack It
The researchers at Linköping University figured out that if the photon source is replaced with a traditional and pulsed light source that floods the interferometers with light, someone can identify the key or the code string. This means it is also possible to read the message without detection. The security test based on Bell’s inequality will not react even though an attack is underway because the photon detectors in the interferometers are swamped by the bright light.
Mohamed Bourennane and two postgraduate students at the Department of Physics at Stockholm University have, in practical experiments, demonstrated that doing things in this way works as an attack. The attacker blinds the detectors with bright light instead of a stream of photons. At the same time, the attacker uses a local hidden variable model, which gives the distribution of the sign and time slot of outcomes for the receivers, to send out incorrect phase information.
This fools the receivers into thinking the system violates Bell’s inequality even though there is no entanglement and no security. The attacker only needs access to the source device to pull this off. The actor can then fully control the key output and break the security of the interferometer system.
The paper suggested solutions to the attack. The researchers listed a number of improved tests and experimental setups that would protect against this. As they put it, “the designer will have to use fast switching and replace the CHSH [Bell] inequality with stronger tests such as modified Pearle-Braunstein-Caves inequalities.”
Quantum computing is still in its infancy. Better practical ways, like those methods suggested by the authors, are needed in order to eliminate the attacks that are possible with the current methods.