Arbitrarily Loss-Tolerant Einstein-Podolsky-Rosen Steering Allowing a Demonstration over 1 km of Optical Fiber with No Detection Loophole

Abstract

Demonstrating nonclassical effects over longer and longer distances is essential for both quantum technology and fundamental science. The main challenge is the loss of photons during propagation, because considering only those cases where photons are detected opens a “detection loophole” in security whenever parties or devices are untrusted. Einstein-Podolsky-Rosen steering is equivalent to an entanglement-verification task in which one party (device) is untrusted. We derive arbitrarily loss-tolerant tests, enabling us to perform a detection-loophole-free demonstration of Einstein-Podolsky-Rosen steering with parties separated by a coiled 1-km-long optical fiber, with a total loss of 8.9 dB (87%).

Alternate abstract:

Plain Language Summary

The world of quantum mechanics has held many surprises for us. One of them is what may be called “spooky action at a distance,” conceptualized first by none other than Einstein, with colleagues Podolsky and Rosen. Imagine two parties, say, Alice and Bob, at some distance apart, sharing a pair of quantum particles that are “entangled”—the two particles are strongly correlated such that together they are in a definite state, even though separately each is in an indefinite state. By making a measurement on her particle, Alice “collapses” Bob’s particle into a definite state and her choice of how to measure her particle affects the set of possible quantum states Bob’s particle is collapsed into. This effect, called “steering” by Schrödinger, does not allow faster-than-light communication but has recently been used to develop new communication protocols. The key practical challenge to scaling these up to long distances is high transmission losses. In this paper, we derive new protocols that, for the first time, enable EPR-steering with arbitrarily high losses, and use them to experimentally demonstrate “spooky action” via an optical fiber over a propagation distance of 1 km.

Experimentally, we implement our protocols by making polarization measurements, at Alice’s and Bob’s stations, on the entangled two-photon states. By evaluating the correlations observed, Bob can use a test—a predetermined correlation function and a bound on it that he calculates—to ensure that EPR-steering has occurred. What gives our protocols their power is (i) the use of many different measurement settings for Alice, and (ii) tailoring of the test to whatever level of loss that applies. Through harnessing both the symmetry of the entangled quantum states and the geometrical properties of our measurement directions, these ideas always allow Bob to distinguish true correlations from artifacts introduced by loss, or even any attempt to cheat on Alice’s part. With arbitrarily many settings our protocols can tolerate an arbitrarily high level of loss. In the experiment we used 16 settings to demonstrate EPR-steering with a total loss of 87%.

Our work opens up new avenues for exploring fundamental quantum phenomena over long distances, as well as performing quantum technology protocols remotely.