Quantifying high-dimensional entanglement with Einstein-Podolsky-Rosen correlations

TL;DR

This paper introduces a scalable method to quantify high-dimensional entanglement using EPR correlations, avoiding complex tomography and providing tight bounds on entanglement measures.

Contribution

It demonstrates that violations of entropic EPR witnesses can directly estimate multiple entanglement measures efficiently in high dimensions.

Findings

Method accurately captures most entanglement in high-dimensional systems

Requires only measurements of correlations between two pairs of observables

Scales efficiently without full quantum tomography

Abstract

Quantifying entanglement in a quantum system generally requires a complete quantum tomography followed by the NP-hard computation of an entanglement monotone --- requirements that rapidly become intractable at higher dimensions. Observing entanglement in large quantum systems has consequently been relegated to witnesses that only verify its existence. In this article, we show that the violation of recent entropic witnesses of the Einstein-Podolsky-Rosen paradox also provides tight lower bounds to multiple entanglement measures, such as the entanglement of formation and the distillable entanglement, among others. Our approach only requires the measurement of correlations between two pairs of complementary observables---not a tomography---so it scales efficiently at high dimension. Despite this, our technique captures almost all the entanglement in common high-dimensional quantum systems, such as spatially or temporally entangled photons from parametric down-conversion.