Entanglement distillation from Gaussian states – Niels Bohr Institute - University of Copenhagen

Niels Bohr Institute > Calendar > 2009 > Entanglement distillat...

Entanglement distillation from Gaussian states

Entanglement distillation is the process of increasing the entanglement between two parties at the expense of the rate of entangled states using only local operations and classical communication. It is essential for combating decoherence in long-distance quantum communication. For discrete variable systems, distillation has been around for several years. For continuous variables (CV), the matter is complicated by the well-known fact that Gaussian operations alone is insufficient to distill entanglement from Gaussian states. Recently, Hage et al. [Nature Physics 4, 915 (2008)], and Dong et al. [Nature Physics 4, 919 (2008)] succeeded in distilling entanglement from CV entangled states that had been subjected to non-Gaussian noise (phase diffusion and time-dependent attenuation). 

Their schemes make use of Gaussian operations such as homodyne detection only. Because of that they are not applicable to linear loss, which is perhaps the most generic type of decoherence and which preserves the Gaussianity of the input state. To address that missing part, we have demonstrated distillation from an initial Gaussian entangled state by using non-Gaussian operations, namely photon subtraction. By probabilistically subtracting a photon at the side of either one or both of the parties (Alice and Bob), we were able to increase an initial weak entanglement as quantified by the logarithmic negativity. Furthermore, in the case of two-photon subtraction, the two-mode squeezing level of the output state was also improved, thereby immediately increasing the state's usefulness for teleportation. The state reconstruction was done by full two-mode homodyne tomography as well as a simplified two-mode tomography method.