30.04.2013 Views

ENTANGLEMENT OF GAUSSIAN STATES Gerardo Adesso

ENTANGLEMENT OF GAUSSIAN STATES Gerardo Adesso

ENTANGLEMENT OF GAUSSIAN STATES Gerardo Adesso

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

94 5. Multimode entanglement under symmetry<br />

The central observation of the present Chapter is embodied by following result<br />

[GA4, GA5], straightforwardly deducible from the discussions in Sec. 2.4.3 and<br />

Sec. 3.1.1.<br />

➢ Unitarily localizable entanglement of bisymmetric Gaussian states. The<br />

bipartite entanglement of bisymmetric (M + N)-mode Gaussian states under<br />

M × N partitions is “unitarily localizable”, namely, through local unitary<br />

(reversible) operations, it can be completely concentrated onto a single pair<br />

of modes, each of them belonging respectively to the M-mode and to the<br />

N-mode blocks.<br />

Hence the multimode block entanglement (i.e. the entanglement between blocks<br />

of modes) of bisymmetric (generally mixed) Gaussian states can be determined as<br />

a two-mode entanglement. The entanglement will be quantified by the logarithmic<br />

negativity in the general instance because the PPT criterion holds, but we will<br />

also show some explicit nontrivial cases in which the entanglement of formation,<br />

Eq. (1.35), between M-mode and N-mode parties can be exactly computed.<br />

We remark that our notion of “localizable entanglement” is different from that<br />

introduced by Verstraete, Popp, and Cirac for spin systems [248]. There, it was<br />

defined as the maximal entanglement concentrable on two chosen spins through<br />

local measurements on all the other spins. 11 Here, the local operations that concentrate<br />

all the multimode entanglement on two modes are unitary and involve the<br />

two chosen modes as well, as parts of the respective blocks.<br />

5.1. Bipartite block entanglement of bisymmetric Gaussian states<br />

In Sec. 2.4.3, the study of the multimode CM σ of Eq. (2.64) has been reduced to<br />

a two-mode problem by means of local unitary operations. This finding allows for<br />

an exhaustive analysis of the bipartite entanglement between the M- and N-mode<br />

blocks of a multimode bisymmetric Gaussian state, resorting to the powerful results<br />

available for two-mode Gaussian states (see Chapter 4). For any multimode Gaussian<br />

state with CM σ, let us define the associated equivalent two-mode Gaussian<br />

state ϱeq, with CM σeq given by<br />

σeq =<br />

<br />

+<br />

ν αM γ ′′<br />

γ ′′T +<br />

ν βN <br />

, (5.1)<br />

where the 2×2 blocks have been implicitly defined in the CM, Eq. (2.65). As already<br />

mentioned, the entanglement of the bisymmetric state with CM σ, originally shared<br />

among all the M + N modes, can be completely concentrated by local unitary<br />

(symplectic) operations on a single pair of modes in the state with CM σeq. Such<br />

an entanglement is, in this sense, localizable.<br />

We now move on to describe some consequences of this result. A first qualificative<br />

remark has been explored in Sec. 3.1.1. Namely, PPT criterion turns out to be<br />

automatically necessary and sufficient for separability of (M + N)-mode Gaussian<br />

states under M × N bipartitions [GA5]. For a more quantitative investigation, the<br />

following symplectic analysis, which completes that of Sec. 2.4.3, will be precious.<br />

11 This (non-unitarily) localizable entanglement will be also computed for (mixed) symmetric<br />

Gaussian states of an arbitrary number of modes, and it will be shown to be in direct quantitative<br />

connection with the optimal fidelity of multiparty teleportation networks [GA9] (see Sec. 12.2 and<br />

Fig. 12.4).

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!