1: %%%%Sample article, book, LateX
2: % Revised An 14.02
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4: \documentclass[a4paper,aps,pra,%twocolumn,
5: %showpacs,
6: preprintnumbers]{revtex4}
7: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
8:
9: %\documentclass[12pt,fleqn]{article}
10: \usepackage{graphics,graphicx,epsfig}
11:
12:
13: %\documentclass[]{article}
14: %\documentstyle[aps,pra,twocolumn,epsfig]{revtex}
15: %\documentclass[fleqn,twocolumn]{article}
16: %\documentclass{book}
17: %\documentstyle[12pt,fleqn,russcorr]{article}% это для ТеХа от
18: %РФФИ.
19:
20: % Это чтобы писать по русски.
21: %\usepackage[cp1251]{inputenc}
22: %\usepackage[russianb]{babel}
23: %\usepackage{mathtext}
24: % Это чтобы писать формулы без проблем
25: \usepackage[centertags]{amsmath}
26: \usepackage{amsfonts}
27: \usepackage{amssymb}
28: %%%%Sample article, book, LateX
29: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30: %\documentclass[twocolumn,fleqn]{article}
31: %\documentclass[fleqn]{article}
32: %\documentclass{book}
33: %\documentstyle[12pt,fleqn,russcorr]{article}% это для ТеХа от РФФИ.
34: %\documentstyle[aps,epsf,draft,12pt]{revtex}
35: %\pagestyle{empty} Это чтобы не было нумерации станиц
36:
37: % Это чтобы писать по русски.
38: %\usepackage[cp1251]{inputenc}
39: %\usepackage[russianb]{babel}
40:
41: % Это чтобы писать формулы без проблем
42:
43: \usepackage[centertags]{amsmath}
44: \usepackage{amsfonts}
45: \usepackage{amssymb}
46: \usepackage{amsthm}
47: \newcommand{\bbbone}{{\mathchoice {\rm 1\mskip-4mu l} {\rm 1\mskip-4mu l}
48: {\rm 1\mskip-4.5mu l} {\rm 1\mskip-5mu l}}}%Эта команда пишет красиво единичный оператор 1.
49: \newcommand{\beq}{\begin{equation}}
50: \newcommand{\eeq}{\end{equation}}
51: \newcommand{\beqa}{\begin{eqnarray}}
52: \newcommand{\eeqa}{\end{eqnarray}}
53:
54: \def\tr{\mathrm{tr}}
55: \newcommand{\ket} [1] {\vert #1 \rangle}
56: \newcommand{\bra} [1] {\langle #1 \vert}
57: \newcommand{\braket}[2]{\langle #1 | #2 \rangle}
58: \newcommand{\proj}[1]{\ket{#1}\bra{#1}}
59: \newcommand{\mean}[1]{\langle #1 \rangle}
60: %\newcommand{\proj}[1]{\ket{#1}\bra{#1}}
61: \newcommand{\kb}[2]{\ket{#1}\bra{#2}}
62:
63: \renewcommand{\Re}{\mathrm{Re}}
64: \renewcommand{\Im}{\mathrm{Im}}
65: \newcommand{\Eqref}[1]{Eq.~(\ref{#1})}
66: \newcommand{\Eqsref}[1]{Eqs.~(\ref{#1})}
67: \newcommand{\Sref}[1]{Sec.~\ref{#1}}
68: \newcommand{\Aref}[1]{Appendix \ref{#1}}
69: \newcommand{\Fref}[1]{Fig.~\ref{#1}}
70:
71: \newcommand{\one}{\mbox{$1 \hspace{-1.0mm} {\bf l}$}}
72:
73: \tolerance = 10000
74:
75: %\numberwithin{equation}{section}
76: %\renewcommand{\theequation}{\thesection.\arabic{equation}}
77: %\numberwithin{equation}{subsection}
78: %\textwidth=17cm \textheight=21cm
79: %\oddsidemargin=-20pt %левое поле=1дюйм
80: %\headsep=-10mm %пробел от колон.тит до текста
81:
82: \begin{document}
83:
84: %\date{}
85: \sloppy
86: \begin{abstract}
87: Large number of multimode entangled states of light generated in
88: down conversion processes belongs to a collection which is natural
89: generalization of the $W$ class. A brief overview of these states,
90: schemes for their preparation, experimental implementations and
91: possible applications are presented.
92: \\
93: \\
94: \\
95: PACS numbers: 03.67.Mn
96: \end{abstract}
97: %\pacs{PACS numbers: 03.67.Mn}
98: \title{On multiparticle $W$ states, their implementations and
99: applications in quantum information problems}
100: \author{V.N. Gorbachev, A.I. Trubilko}.
101: \affiliation {Laboratory for quantum information $\&$
102: computation,
103: University of AeroSpase Instrumentation,\\
104: St-Petersburg, 190000, Bolshaia Morskaia 67, Russia.}
105:
106: %\begin{document}
107: \maketitle
108: %\tableofcontents
109: \section{Introduction}
110:
111: Entangled states have been generated in many experiments and
112: their applications in quantum communications have been
113: demonstrated (see for example \cite{EntExp}). Now great efforts
114: are concentrated on investigation of multiparticle entanglement
115: with its promising features which are interesting in
116: decoherence-free quantum informational processing, advanced
117: multiparty quantum communications and others.
118:
119: When considering implementation of the multiparticle
120: entangled states (MES), one finds that most of them have been
121: generated in optics experiments. In a typical optical scheme
122: photons of the source are distributed to output modes by linear
123: lossless elements. If the configuration is symmetric and a photon
124: enters input, then it can be found on any of the outputs. By this
125: way the multimode light is achieved, its state belongs to the
126: $W$-class introduced by Cirac et al \cite{Cirac W}. Varying
127: configuration of the scheme as well as using the source of polarized
128: photons a more extensive collection of MES than $W$-class arises.
129: Particularly it includes the Dicke states and can be studied from
130: point of view of quantum information theory (QIT)
131: % Term QIT from Werner
132: without referring to any physical system. Analysis of common
133: features gives us an answer how to manipulate MES in optimal way
134: and which of informational tasks can be done using a given
135: entangled state. In respect to physics MES are natural states of
136: the multiparticle systems. So if any $m$ particles of a large
137: ensemble can be excited then all possibilities result in the
138: considered MES. Indeed entangled states of two macroscopic atomic
139: ensembles have been demonstrated experimentally by Polzik et al
140: \cite{Polzik Nature}.
141:
142: It is important to know the common features of MES because it
143: determines how to detect them in experiment. In QIT there is a set
144: of criteria of entanglement which require the knowledge of the
145: state or its entropy also Bell inequalities are often discussed.
146: But exploiting such criteria in experiment is a hard problem.
147: However there are specific witness observables which
148: expectation values indicate entanglement. Experimental
149: implementation of the witness observable for polarized light have
150: been demonstrated by Weinfurter et all \cite{Weinfurter Witness}.
151: The estimation of an unknown state can be made by quantum
152: tomography. Indeed this method has been used by Roos et all
153: \cite{IonExpW} in experiment with trapped ions. Another way is
154: measuring a such operator which eigenvector is the desired
155: entangled state. Then its eigenvalue and variance indicate
156: entanglement. In fact some members of the $W$-class are reduced
157: to the Dicke states which are eigenvectors of two collective
158: operators $J^{2}$ and $J_{3}$ \cite{Dicke}. In optics
159: implementation the spin variables can be associated with
160: polarization of light and one can measure, for example, variances
161: of these operators which describe noise of light. For eigenvectors
162: these variances are equal to zero and it means that there is no
163: noise. More precisely in such measurement the shot noise of light
164: is suppressed bellow standard quantum limit given by the coherent
165: state.
166:
167: Quantum correlations
168: are fragile and easily destroyed with environment nevertheless
169: several MES have immunity to decoherence and they are robust to
170: loss of particles. These features are attractive for quantum
171: communications, but their exploiting is a hard problem. For
172: example, many attempts have been made to introduce $W$ state
173: instead of the EPR pair in the standard teleportation protocol
174: proposed by Bennett et al \cite{Bennett tel}. But most of these
175: proposals results in conditional teleportation, when the task is
176: accomplished with a probability.
177: %\footnote{Conditional tel with W}.
178: However several of the $W$-states can be suitable as a quantum
179: channel for dense coding and the unconditional teleportation of
180: entangled states
181: %\footnote{Our W channel}
182: also for the problem of secrete sharing
183: %\footnote{Secret sharing}
184: and other tasks.
185:
186: The main aim of this work is to consider properties,
187: implementations and applications of the collection of MES, which
188: are simple generalization of the $W$ class. For a particular case
189: of three-particle $GHZ$ and $W$ states transformations between
190: them, three-party quantum communications protocols for secrete
191: sharing and splitting of quantum information with $GHZ$ have been
192: discussed by Karlsson et al \cite{Karlsson Rew 3W 3GHZ}.
193: %\\
194: %\footnote{A. Karlsson, M.Bourennane, I.Ghiu, D. Ljunggren, A.
195: %Manson. Some Properties of Three-party Entangled States and their
196: %Application in Quantum Communication. Quantum Computers and
197: %Computing, V.4, p.3 (2003)}.
198: %\\
199: In our work using the standard approaches of QIT we pay attention
200: to physical features of MES.
201:
202: This paper is organized as follows. First we describe
203: tree-particle $W$ state, which is non-equivalent to $GHZ$ and
204: robust to loss of particle. Then more general states are
205: introduced and their connection with the Dicke states together
206: with their properties, measures of entanglement and witness
207: observables are considered. Next we examine several proposals for
208: generating MES of atoms and light and overview experimental
209: implementations of the three and four photon $W$ states. Finally
210: several protocols are discussed.
211:
212: \section{Properties}
213: \subsection{Three-qubit GHZ and W states}
214:
215: \subsubsection{Classification of states in LOCC.}
216: %\\
217: There are strong definitions for the three-particle $W$ states
218: introduced by Cirac et al \cite{Cirac W}. These definitions are
219: based on a set of transformations known as LOCC (Local Operations
220: and Quantum Communication). We will use a simple notation, that
221: operators are local if $U\ket{\varphi}_{AB}=A\otimes
222: B\ket{\varphi}_{AB}$, where operator $A$ acts on the particle $A$
223: and don't affect to $B$ and so on. Following this point one finds,
224: that the two-mode hamiltonian described a parametric down
225: conversion source (PDC) $H=ik\hbar(a^{\dagger}b^{\dagger}-ab)$ is
226: an example of non-local operator. PDC generates entangled states
227: but they can't be created by LOCC. This is a general property of
228: entanglement.
229: \\
230: How to compare one state with another? From physical reasons the
231: answer is clear. If a physical system is prepared in different
232: states, then by measuring of observables, one can distinct them in
233: principle. In QIT, which operates the logical states without
234: referring to any particular physical system, two states are
235: identical, if they can be obtained from each other with certainty
236: by LOCC.
237: %\footnote{For general case it needs introduce equivalence
238: %under stochastic LOCC, when two states can be converted from one
239: %to another by LOCC with some probability.}.
240: Particularly, it means, that parties can use these two states for
241: the same task \cite{C.H. Bennett9511030}.
242: \\
243: In this approach there are two classes of irreducible tripartite
244: entanglement, namely either GHZ states \cite{Greenberger Bell's
245: theorem}
246: \begin{eqnarray}
247: % \nonumber to remove numbering (before each equation)
248: \ket{GHZ}=1/\sqrt{2}(\ket{000}+\ket{111}),
249: \end{eqnarray}
250: and
251: \begin{eqnarray}
252: \label{W}
253: % \nonumber to remove numbering (before each equation)
254: \ket{W}=1/\sqrt{3}(\ket{100}+\ket{010}+\ket{001}),
255: \end{eqnarray}
256: known as W-state in QIT. Here $\ket{0},\ket{1}$ are states of a
257: two-level system, or qubit (quantum bit). One finds, that
258: $\ket{GHZ}$ can't be converted into $\ket{W}$ by LOCC, but they
259: are entangled because of their wave functions are not factorized
260: into product of three particles. For general case the $W$-class
261: introduced by Cirac has the form
262: $\varphi_{W}=a\ket{000}+b\ket{100}+c\ket{010}+d\ket{001}$, where
263: $a^{2}+b^{2}+c^{2}+d^{2}=1$.
264:
265: \subsubsection{Robustness}
266:
267: Some differences between $\ket{GHZ}$ and $\ket{W}$ are clear
268: without LOCC.
269: \\
270: Consider any two of three particles, say 1 and 2, which density
271: matrix is $\rho(12)=Tr_{3}\rho(123)$. If one of the particles is
272: traced out, in QIT it means a loss of particle. In fact three
273: parties share particles 1,2 and 3 in entangled state and one of
274: them decides not to cooperate with other two. Can the remainder
275: two parties accomplish the task? Answer depends on the robustness
276: of the state. For $GHZ$
277: \begin{eqnarray}
278: % \nonumber to remove numbering (before each equation)
279: \rho_{GHZ}(12)=1/2(\proj{00}+\proj{11}),
280: \end{eqnarray}
281: for $W$
282: \begin{eqnarray}
283: \label{RRR}
284: % \nonumber to remove numbering (before each equation)
285: \rho_{W}(12)=1/3\proj{00}+2/3\proj{\Psi^{\dagger}},
286: \end{eqnarray}
287: where $\Psi^{\dagger}=1/\sqrt{2}(\ket{01}+\ket{10})$. The main
288: difference is that the density matrix $\rho_{W}(12)$ has
289: non-diagonal elements or coherence $\ket{01}\bra{10}$. This is a
290: reason of entanglement of $\rho_{W}(12)$ in contrast to
291: $\rho_{GHZ}(12)$.
292: \\
293: In more detail. To analyze the mixed states it needs to introduce
294: criteria of inseparability, that are generalizations of the
295: non-factorizability of the wave function. A criterion of Werner \cite{Werner} tells, that
296: state is separable or classically correlated or non-entangled
297: if its density matrix has the form
298: \begin{eqnarray}
299: % \nonumber to remove numbering (before each equation)
300: \rho(12)=\sum_{k}\lambda_{k}\rho(1)_{k}\otimes\rho_{k}(2),
301: \end{eqnarray}
302: where $\sum_{k}\lambda_{k}=1$ and all $\lambda_{k}\in [0,1]$. Then
303: one finds that $\rho_{GHZ}(12)$ is classically correlated. Next
304: criterion is necessary and sufficient to establish inseparability
305: of $\rho_{W}(12)$. It tells, that the state which dimension of
306: Hilbert space is $2\times 2$ or $2\times 3$ is inseparable, if any
307: of eigenvalues of the partially transposed density matrix is
308: negative \cite{Peres}, \cite{Horodecki Criterium}. The partially
309: transposed density matrix, say over particle 1, reads
310: $\rho^{T_{1}}(12)_{W}=1/3\proj{00}+2/3(\proj{01}+\proj{10}+
311: \ket{11}\bra{00}+\ket{00}\bra{11})$, it has a negative eigenvalue,
312: then $\rho_{W}(12)$ is inseparable or entangled. In this case
313: quantum correlations between the remainder two particles survive
314: after tracing and this is robustness to particle loss.
315: \\
316: \subsection{Several generalizations}
317:
318: \subsubsection{Multiparticle $W$, $ZSA$ and other states}
319:
320:
321: A simple generalizations is possible by introducing the
322: n particle states of the form
323: \begin{eqnarray}
324: \label{eta}
325: % \nonumber to remove numbering (before each equation)
326: \eta_{n}(1)=q_{1}\ket{10,\dots,0}+q_{2}\ket{01,\dots,0}+\dots
327: +q_{n}\ket{00,\dots,1}),
328: \end{eqnarray}
329: where $\sum_{k}|q_{k}|^{2}=1$. In particular case
330: \begin{eqnarray}
331: % \nonumber to remove numbering (before each equation)
332: q_{1}+\dots q_{n}=0,
333: \end{eqnarray}
334: one finds the zero sum amplitude (ZSA) states introduced by Pati
335: \cite{Pati}. They are not equivalent to $GHZ$ under LOCC. If all
336: coefficients in $\eta_{n}(1)$ are equal there is a totally
337: symmetric wave function
338: \begin{eqnarray}
339: % \nonumber to remove numbering (before each equation)
340: W_{n}=1/\sqrt{n}(\ket{10,\dots,0}+\ket{01,\dots,0}+\dots
341: +\ket{00,\dots,1}),
342: \end{eqnarray}
343: that is known in QIT as multiparticle W-states \cite{Cirac W}.
344: Such symmetric vector describes an ensemble of two-level physical
345: systems, qubits, where only $m=1$ particle from $n$ is excited.
346: When $m\geq 1$ the symmetrized states has the form
347: \begin{eqnarray}
348: \label{Sym}
349: % \nonumber to remove numbering (before each equation)
350: \ket{m;n}=1/Q\sum_{z}P_{z}\ket{\underbrace{1,\dots
351: 1}_{m},\underbrace{0,\dots,0}_{n-m}},
352: \end{eqnarray}
353: where $P_{z}$ is one from $C^{n}_{m}=n!/(m!(n-m)!)$
354: distinguishable permutations of particles, $Q=\sqrt{C^{n}_{m}}$.
355: All states in the superposition (\ref{Sym}) have equal weight
356: $Q$, when their weights are different the more general
357: $\eta_{mn}$ states may be found, which are a natural
358: generalization of the $W$-class. The introduced collection of
359: $\eta_{mn}$ contains $W$, $ZSA$ and symmetric states $\ket{m;n}$.
360: \\
361: These states seem to be natural for multiparticle systems and can
362: be generated in many physical processes. When considering problem
363: of interaction between atoms and light, for example, the
364: collective atomic operators are introduced. For two-level
365: identical atoms they read
366: \begin{eqnarray}
367: % \nonumber to remove numbering (before each equation)
368: \label{atS}
369: S_{xy}=\sum_{a}\ket{x}_{a}\bra{y},
370: \end{eqnarray}
371: where $\ket{x}_{a}\bra{y}, x,y=0,1$ is operator of a single atom
372: $a$ and 0,1 label lower and upper level. There is a representation
373: for $\ket{m;n}$
374: \begin{eqnarray}
375: % \nonumber to remove numbering (before each equation)
376: S_{10}^{m}\ket{0,\dots,0}=m!Q\ket{m;n}.
377: \end{eqnarray}
378: It tells, that any symmetric states $\ket{n;m}$, particulary $W$
379: ones, can be produced in any process of collective interaction
380: between atoms and light or other system. There is a simple
381: physical reason for it. When each of $m\leq n$ identical atoms
382: absorb a photon, then all possibilities results in superposition
383: $\ket{m;n}$, when atoms are distinguished more general state
384: $\eta_{nm}$ arises.
385:
386: \subsubsection{Connection with Dicke states}
387: Some members of the $W$ class can be reduced to the Dicke states.
388: Let introduce collective operators $J_{k}, k=1,2,3$ and
389: $J^{2}=J_{1}^{2}+J_{2}^{2}+J_{3}^{2}$, that obey the commutation
390: relations of the momentum operators
391: \begin{eqnarray}
392: % \nonumber to remove numbering (before each equation)
393: [J_{j};J_{k}]=i\epsilon_{jkl}J_{l}.
394: \end{eqnarray}
395: The Dicke states are defined as eigenvectors of two operators
396: $J^{2}$ and $J_{3}$ \cite{Dicke}
397: \begin{eqnarray}
398: %\nonumber to remove numbering (before each equation)
399: J^{2}\ket{jl}=j(j+1)\ket{jl},~~~J_{3}\ket{jl}=l\ket{jl},
400: \end{eqnarray}
401: where $|l|\leq j$, $\max j=n/2$, $n$ is number of the particles
402: and $l=[(n-m)-m)]/2$ is difference between the numbers of
403: non-excited and excited particles. Using (\ref{atS}), we have a
404: representation
405: \begin{eqnarray}
406: %\nonumber
407: \label{rrr}
408: J_{1}=(S_{10}+S_{01})/2,~~J_{2}=i(S_{10}-S_{01})/2,~~J_{3}=(S_{00}-S_{11})/2.
409: \end{eqnarray}
410: Note, that in (\ref{rrr}) the vectors $\ket{0},\ket{1}$ however
411: can be considered as the Fock state of light.
412: \\
413: All symmetric states have the form $\ket{j=n/2,l}$ and
414: \begin{eqnarray}
415: % \nonumber to remove numbering (before each equation)
416: W_{n}=\ket{j=n/2,~l=n/2-1}.
417: \end{eqnarray}
418: For ZSA states
419: \begin{eqnarray}
420: % \nonumber to remove numbering (before each equation)
421: \eta_{n}=\ket{j=n/2-1,~ l=n/2-1}.
422: \end{eqnarray}
423: As result $W$ and $ZSA$ are eigenvectors of $J^{2}$ and $J_{3}$
424: and belong to family of the Dicke states.
425:
426: \subsubsection{Symmetry and Decoherence-Free states}
427:
428: When the symmetry of state is conserved in a physical processes
429: one finds, for example, that the antisymmetric two-particle wave
430: function $\Psi^{-}=(1/\sqrt{2})(\ket{01}-\ket{10})$ can't be
431: transformed into product $\ket{00}$.
432: Such entanglement is therefore
433: decoherence-free (DF). A state is DF if it is invariant under some
434: unitary transformation, described an collective interaction with
435: noisy environment. It is interesting for protection of information
436: by the noiseless quantum code, that has been considered by Zanardi
437: et al \cite{Zanardi Noiseless}.
438: \\
439: Let $\Psi^{-}$ be the state of two atoms, which interact with its
440: thermostat, then
441: \begin{eqnarray}
442: % \nonumber to remove numbering (before each equation)
443: \label{dec}
444: S_{01}\Psi^{-}=0.
445: \end{eqnarray}
446: Formally $\Psi^{-}$ looks as a vacuum and has immunity to the
447: spontaneous decay. This state can be storage in a collective
448: thermostat which has been considered by Basharov \cite{Askhat}. It
449: is described by equation of the Lindblad form\begin{eqnarray}
450: \dot{\rho}=-\gamma[R^{\dagger}R\rho-R\rho R^{\dagger}+h.c.],
451: \end{eqnarray}
452: where $\gamma$ is a decay rate and $R=S_{01}$ is a collective
453: operator.
454: \\
455: To preserve entanglement, in QIT a class of DF states has been
456: introduced \cite{Palma}, \cite{Zanardi Noiseless}, \cite{Lidar},
457: \cite{Kempe}. For two particles there is only one DF state
458: $\Psi^{-}$. In the case of four particles there are two DF states,
459: which are interesting for applications. They are a product of
460: singlets $\Psi^{-}$
461: \begin{eqnarray}
462: \label{DF1}
463: \Phi_{0}=\ket{\Psi^{-}}\ket{\Psi^{-}}&&
464: \end{eqnarray}
465: and an orthogonal to $\Phi_{0}$ vector, introduced by Kempe
466: \cite{Kempe}
467: \begin{eqnarray}
468: \label{DF2}
469: \Psi_{1}=(1/\sqrt{3})(\ket{0011}+\ket{1100}-
470: \ket{\Psi^{+}}\ket{\Psi^{+}}).
471: \end{eqnarray}
472: To protect quantum information, the logical qubit
473: $\alpha\ket{0}+\beta\ket{1}$ can be encoded in to superposition
474: $\alpha\Psi_{0}+\beta\Psi_{1}$, which is DF state and immune
475: against noise. These two DF states have been generated
476: experimentally by Weifurter et al \cite{Weifurter DF} to
477: demonstrate DF quantum information processing.
478: \\
479: Several examples of DF states of physical systems can be
480: introduced by a simple generalization of (\ref{dec}). Consider
481: $\eta_{n}(1)$, where logical qubits are implemented by two-level
482: atoms or by modes of light in the Fock state with $0$ and $1$
483: photon. Next observation is true. There is a collective operator
484: $R$, for which
485: \begin{eqnarray}
486: % \nonumber to remove numbering (before each equation)
487: \label{decN} R\eta_{n}=\sum_{k}q_{k}\ket{0}.
488: \end{eqnarray}
489: It can be chosen in the form $R=S_{10}$ or $R=\sum_{k}a_{k}$,
490: where $a_{k}$ is annihilation operator of the light mode $k$. If
491: $\sum_{k}q_{k}=0$, one finds ZSA states, which are robust to
492: Decoherence like $\Psi^{-}$. So that either atomic ensemble or
493: light is prepared in ZSA state it may conserve its quantum
494: correlations under a collective noisy environment.
495:
496: \subsection{Entanglement of multiparticle $W$ states}
497:
498: \subsubsection{Entanglement and its measures}
499:
500: Entanglement of multiparticle system can depend from the number of
501: particles. The reason is that if single excitation is distributed
502: into a large number of particles then total state is closer to
503: unexcited or vacuum state.
504: \\
505: In more detail. Introducing a density matrix of $n$ particles
506: $\rho(n)= \proj{W_{n}}$ and considering a state of any $s\leq n$
507: particles we have
508: \begin{eqnarray}
509: % \nonumber to remove numbering (before each equation)
510: \label{rho s}
511: \rho(s\leq n)=(s/n)\proj{W_{s}}+(1-s/n)\proj{0}.
512: \end{eqnarray}
513: For $s=2$ the Peres-Horodecki inseparability criterion can be
514: applied. The eigenvalues of the partially transposed density
515: matrix are $\lambda=\{1/n;1/n;(n-2)[1\pm\sqrt{1+4/(n-2)^{2}}]/2n$,
516: where one of them is negative. Thus the state is inseparable or
517: entangled. However in the limit of $n\to\infty$ entanglement
518: vanishes: $\lambda=\{0;0;0;1\}$, that is in agreement with
519: (\ref{rho s}), from which it follows, that $\rho(s\leq
520: n)\approx\proj{0}$.
521: \\
522: The problem whether a given multiparticle state is entangled is
523: hard because of the calculation difficulty increases exponentially
524: with number of particles.
525: There is
526: no necessary and sufficient operational criterion and various
527: measures of entanglement are used. Several common measures are
528: entanglement entropy, entanglement formation and negativity.
529: \\
530: Entanglement entropy $E(\ket{\Psi}_{AB})$ of a pure state and a
531: partition for the system $A,B$ is defined as
532: $E(\ket{\Psi}_{AB})=S(\rho_{A})=S(\rho_{B})$, where
533: $S(\rho)=-Tr(\rho\log \rho)$ is von Neuman entropy and
534: $\rho_{A}=Tr_{B}\proj{\Psi_{AB}}$ is the reduced density matrix.
535: For product states entanglement entropy is zero. It has its
536: maximum $\log dim(A)$, given for a partition with dimension
537: $dim(A)=d_{A},dim(B)=d_{B}$ and $d_{A}<d_{B}$. A state that
538: achieves this maximum is maximally entangled:
539: $\Psi_{AB}=1/\sqrt{d_{A}}(\ket{0,0}+\ket{1,1}+
540: \dots+\ket{d_{A}-1,d_{A}-1})_{AB}$. For EPR pair of the form
541: $\ket{\varphi}=\alpha\ket{00}+\beta\ket{11}$ we have
542: $E(\ket{\varphi})=H(p)$, where $H(p)=-p\log p-(1-p)\log (1-p)$ is
543: a function of entropy, well known in the theory of information,
544: $p=|\alpha|^{2}=1-|\beta|^{2}$. If $p=1/2$ the function $H(p)$ has
545: its maximum corresponding to maximal entanglement. Using the
546: presented definition the entanglement entropy of the $W$-state
547: obtained from (\ref{rho s}) has the form of $H(p)$:
548: \begin{eqnarray}
549: % \nonumber to remove numbering (before each equation)
550: E(\ket{W_{n}})=-(s/n)\log (s/n)-(1-s/n)\log (1-s/n),
551: \end{eqnarray}
552: where $A$ and $B$ are subsystems of $s$ and $n-s$ particles. If
553: $s=n/2$, then $A$ and $B$ have the same number of particles and
554: their entanglement achieves maximum.
555: \\
556: Entanglement of formation is reduced to entanglement entropy for
557: pure states and defined as
558: $E_{F}(\rho_{AB})=\min_{\{p_{i}\ket{\psi_{i}}_{AB}\}}
559: \sum_{i}p_{i}E(\ket{\psi_{i}}_{AB})$, where
560: $\rho_{AB}=\sum_{i}p_{i}\proj{\psi_{i}}$.
561: \\
562: The logarithmic negativity is defined as the absolute sum of the
563: negative eigenvalues of the partial transpose with respect to $A$
564: of density matrix $\rho_{AB}$. So
565: $N(\rho_{AB})=\sum_{i}(|\lambda_{i}|-\lambda_{i})/2$. Negativity
566: may also disagree with other measures for the so-called positive
567: partial transposed entangled states, which negativity is zero. It
568: demonstrates, that to analyze multiparticle entanglement it needs
569: various measures and criteria.
570: \\
571: There is a criterion, that can be interesting from the
572: experimental point of view. It is based on single-particle
573: measurement. The persistency of entanglement is defined as the
574: minimum number of single-particle measurements $M$ such that, for
575: all measurement outcomes, the state is completely disentangled
576: (separable) \cite{BriegelRaussendorf}. Let $\ket{0},\ket{1}$ be a
577: basis of the measurement. Then for GHZ $M=1$, but for $W_{n}$ it
578: needs $M=n-1$ measurements to obtain a separable state.
579:
580: \subsubsection{Witness}
581:
582: In practice exploiting of the above measures is a hard problem and
583: any recipes adjusted to observables seem to be more appropriate.
584: \\
585: To detect various forms of multipatite correlations witness have
586: been introduced \cite{Horodecki Criterium}, \cite{Terhal}. A
587: witness of $n$- particle entanglement is an observable, which
588: value on state with $n-1$ partite entanglement is positive and
589: negative on some $n$-partite entangled state.
590: \\
591: A witness operator for the three-particle $W$ state reads
592: \begin{eqnarray}
593: % \nonumber to remove numbering (before each equation)
594: %\nonumber
595: W^{(1)}_{W}=2/3-\proj{W}.&&
596: %W^{(2)}_{W}=1/2-\proj{GHZ'}
597: \end{eqnarray}
598: In accordance with definition $Tr[W^{(1)}_{W}\proj{W}]=-1/3$ and
599: $Tr[W^{(1)}_{W}\rho_{W}(12)]=1/9$, where two particle density
600: matrix $\rho_{W}(12)$ is given by (\ref{RRR}). This witness has
601: positive expectation value on biseparable and fully separable
602: states. Then it detects all tripartite entangled states of $W$ and
603: $GHZ$ classes, which can be distinguished by the second witness
604: operator
605: \begin{eqnarray}
606: % \nonumber to remove numbering (before each equation)
607: %\nonumber
608: W^{(2)}_{W}=1/2-\proj{GHZ'},
609: \end{eqnarray}
610: where $\ket{GHZ'}$ is obtained from $\ket{GHZ}$ by replacing
611: $\ket{x}\to ((-1)^{x}\ket{0}+i\ket{1})/\sqrt{2}$, $x=0,1$.
612: \\
613: These witnesses can be observed in experiment. Using the Pauli
614: matrixes one finds
615: \begin{eqnarray}
616: \label{Ww1}
617: \nonumber
618: W^{(1)}_{W}= 1/24[17+7\sigma_{z}^{\otimes
619: 3}+3(\sigma_{z}\otimes1\otimes 1 +1\otimes \sigma_{z}\otimes
620: 1+1\otimes1\otimes\sigma_{z})+5 (\sigma_{z}^{\otimes 2}\otimes
621: 1 +\sigma_{z}\otimes 1\otimes\sigma_{z}+1\otimes\sigma_{z}^{\otimes 2})&&\\
622: \nonumber -(1+\sigma_{z}+\sigma_{x})^{\otimes 3}-
623: (1+\sigma_{z}-\sigma_{x})^{\otimes 3}-
624: (1+\sigma_{z}+\sigma_{y})^{\otimes 3}
625: -(1+\sigma_{z}-\sigma_{y})^{\otimes 3}].&&
626: \end{eqnarray}
627: In optics implementations two qubit states can be presented by the
628: polarization of the photons with horizontal $H$ and vertical $V$
629: linear polarization. Then it needs a set of polarization analyzers
630: to measure the witness operators \cite{Weifurter Witness}. Being
631: non-universal measure, witnesses provide the sufficient criteria.
632: \\
633: Another way of testing the $W$ entanglement is to measure the
634: operators $J^{2}$ and $J_{3}$ whose eigenvectors are $W$ and $ZSA$
635: states. In the representation given by (\ref{rrr}) the spin
636: variables can be associated with the $H$ and $V$ polarized
637: photons. Then by measuring the polarization of photons one can get
638: the expectation values of $J^{2}$ and $J_{3}$, whose variances are
639: zero for $W$ and $ZSA$ states. These expectation values and its
640: variances indicate the entanglement.
641: \section{Schemes for generation.}
642:
643: \subsection{Atomic systems}
644: \subsubsection{Schemes}
645: There are several proposals on generating of $W$ states in Cavity
646: Quantum Electrodynamics (QED) and Raman interaction between
647: three-level atoms and the high-Q cavity modes. These models seem
648: to be attractive but they often neglect all relaxations processes
649: that is a hard problem in its implementation. Some principal
650: features can be demonstrated by considering a more simple model of
651: two-level atoms in free space.
652: \\
653: Interaction between an ensemble of two-level atoms and light can
654: be described by the usual Hamiltonian
655: \begin{eqnarray}
656: \label{hHH} H=i\hbar (S_{10}B-S_{01}B^{\dagger}),
657: \end{eqnarray}
658: where $B$ is a field operator. This Hamiltonian allows to examine
659: various processes of the one-photon interaction, for which $B=ga$,
660: Raman type scattering of two modes $a$ and $b$, when
661: $B=fa^{\dagger}b$, where $g,f$ are coupling constants. In the
662: model given by (\ref{hHH}) there are some physical reasons, that
663: result in $W$ states. Without relaxation one finds integral of
664: motion conserving the total number of excitations $m$. For
665: example, if all atoms are in their ground states and light in the
666: Fock state with one photon only, then $m=1$. In the case of
667: one-photon interaction the integral has the form
668: %\begin{eqnarray}
669: % \nonumber to remove numbering (before each equation)
670: $I=a^{\dagger}a-(1/2)[S_{00}-S_{11}].$
671: %\end{eqnarray}
672: Then during evolution atoms and light exchange the excitation. As
673: result $m$ is distributed into atomic ensemble or into the light
674: modes and the symmetric Dicke states $\ket{m,n}$ particularly
675: $W_{n}$ states of either atoms or modes are generated.
676: \\
677: Generally the multiparticle model given by (\ref{hHH}) is not
678: integrable. But for particular cases simple exact solutions can be
679: found. The reason is that in the symmetry-preserving interaction
680: only a part of states from the total Hilbert space is involved in
681: evolution. Then one can get simple analytic solutions, if the
682: problem includes a small number of $m$. In the case of Raman type
683: interaction we have \cite{Gorbachev parametric type interaction}
684: \begin{eqnarray}
685: \label{RI} \nonumber
686: (\alpha\ket{01}+\beta\ket{10})_{ab}\otimes\ket{m;n}\to &&\\
687: \nonumber
688: \alpha\{\cos\theta_{m}\ket{01}\otimes\ket{m;n}+
689: \sin\theta_{m}
690: \ket{10}\otimes\ket{m+1;n}
691: \}&&\\
692: +\beta\{-\sin\theta'_{m}\ket{01}\otimes\ket{m-1;n}+\cos
693: \theta'_{m}\ket{10}\otimes\ket{m;n}\},
694: \end{eqnarray}
695: where $|\alpha|^{2}+|\beta|^{2}=1$,
696: $\theta_{m}=tf\sqrt{(m+1)(n-m)}$, $\theta'_{m}=\theta_{m-1}$. This
697: example shows evolution of the totally symmetric initial state of
698: atoms $\ket{m;n}$ and entangled state
699: $\alpha\ket{01}+\beta\ket{10}$ of two modes. Equation (\ref{RI})
700: describes the next processes: 1/ generation of atomic $W$
701: entanglement $\ket{0;n}\to\ket{1;n}=W_{n}$, 2/ transformation of
702: the symmetric states $\ket{m;n}\to\ket{m\pm 1;n}$, 3/ entanglement
703: swapping, when the light state is transformed into atoms and back.
704: \\
705: Preparation of the $W$ and $GHZ$ atomic states in Raman type
706: interaction has been considered by Agarwal et al \cite{Agarwal}.
707: In this work the numerical analyze of the analytic solutions has
708: been presented.
709: \\
710: If an excited atom interacts with three or more cavity modes, it
711: can emits a photon into one of them, then the $W$ state of light
712: may be achieved \cite{Zhuo-Liang Cao W state via cavity QED}.
713: \\
714: Measurement is another way for preparation of a physical system
715: in a given state, but it can be done with some probability. If
716: atoms interact with cavity modes, then by detecting an output
717: photon the atomic $W$ and Dicke states may be achieved
718: \cite{Yun-Feng Xiao}.
719: \\
720: Interaction between atoms and light can produce entanglement
721: between them. Such systems are useful for preparing entangled
722: state of atomic ensembles, when a projective measurement on
723: photons is performed.
724: By this way $W$ states of atomic ensembles can be achieved
725: \cite{Gorbachev entangled W-states atom}, they have a hierarchic
726: organization being consisting of ensembles each of which is in the
727: $W$ state.
728: \\
729: Indeed the Heisenberg model was used to produce three-atom or
730: four-atom $W$ state in Ref. \cite{X. Wang}.
731: \subsubsection{Experiment with trapped ions}
732: Three qubit W and GHZ states of trapped ions have been generated
733: experimentally by Roos et all \cite{IonExpW} and conditional
734: operations for readout of an individual qubit have been
735: implemented.\\
736: In this experiment qubits are encoded in the ground and metastable
737: states $D$ and $S$ of the $^{40}Ca^{+}$ ion. A laser pulse can
738: rotate each ion
739: \begin{eqnarray}
740: R(\theta,\phi)=\exp[i\theta/2(e^{i\phi}\sigma^{+}+e^{-i\phi}\sigma^{-})],
741: \end{eqnarray}
742: that results in transitions between levels $D$ and $S$, where
743: $\sigma^{+}=\ket{S}\bra{D}$. By this way the $\pi/2$ pulse, for
744: which $R(\pi/2,0)$, creates a superposition
745: $1/\sqrt{2}(\ket{S}+i\ket{D})$, if initially ion is in $S$ state.
746: When ions are trapped in a linear Pauli trap each of them can
747: interact with vibrational mode due from motion
748: \begin{eqnarray}
749: R^{+}(\theta,\phi)=\exp[i\theta/2(e^{i\phi}\sigma^{+}b^{\dagger}+e^{-i\phi}\sigma^{-}b)],
750: \end{eqnarray}
751: where $b,b^{\dagger}$ are photon operators of the mode. Both
752: operations $R$ and $R^{+}$ were implemented experimentally. Using
753: them one can entangle ions with vibrational mode, rotate each ion,
754: map the state of modes into ions and other. As result
755: from the initial state of the trapped
756: ions $\ket{SSS}$ the desired entanglement can be prepared.
757: \\
758: Indeed evolution given by $R$ and $R^{\dagger}$ has the same form,
759: when only one photon of vibrational mode is involved. Let
760: $\ket{0}=\ket{S}$ or $\ket{S}\otimes\ket{1}_{b}$ and
761: $\ket{1}=\ket{D}$ or $\ket{D}\otimes\ket{0}_{b}$, then one finds
762: \begin{eqnarray}
763: %\nonumber
764: R,R^{\dagger}:~~\alpha\ket{0}+\beta\ket{1}\to
765: (\alpha\ket{0}+\beta\ket{1})\cos\theta/2%&&\\
766: +i (\alpha e^{-i\phi}\ket{1}+\beta
767: e^{i\phi}\ket{0})\sin\theta/2,%&&
768: \end{eqnarray}
769: where $|\alpha|^{2}+|\beta|^{2}=1$.
770: \\
771: To generate W state it needs a sequence of 5 laser pulses
772: addressed to ion 2, 3 and 1
773: \begin{eqnarray*}
774: R^{+}_{2}(2
775: \arccos(1/\sqrt{3}),0)R_{3}(\pi,\pi)R_{3}^{+}(\pi/2,\pi)%&&\\
776: R_{1}(\pi,0)R_{1} ^{+}(\pi,\pi),&&
777: \end{eqnarray*}
778: where first $R^{+}_{2}(2 \arccos(1/\sqrt{3}),0)$ is a
779: beamsplitter-like pulse on ion 2, which entangles its state with
780: the vibrational mode generating a non-symmetric superposition
781: $1/\sqrt{3}(\ket{SSS}\ket{0}_{b}+i\sqrt{2}\ket{SDS}\ket{1}_{b})$.
782: Next pulses result in the W state of ions
783: $1/\sqrt{3}(\ket{DDS}+\ket{DSD}+\ket{SDD}).$ In this experiment
784: for reconstruction of density matrix the state tomography has been
785: used, and fidelity of 83$\%$ was observed.
786: \\
787: The $^{40}Ca^{+}$ ion has an additional Zeeman level $D'$ so that
788: laser pulse on $S-D'$ transition can map the state
789: $\ket{S}\to\ket{D'}$ and back. The mapping allows readout
790: individual ion from the quantum ion trapped register while
791: preserving coherence. In the experiment with three-partite
792: entanglement the states of two ions were mapped into $S-D'$ space.
793: After reading the remainder ion, the laser pulses remap the states
794: into original space preserving coherence.
795: \\
796: The presented technics allows generating and manipulating
797: entanglement and are promising for quantum computing.
798:
799:
800: \subsection{Optical schemes for light}
801:
802: \subsubsection{States of light and structure of the schemes}
803:
804: For quantum state engineering the optical implementation
805: of the $W$ states is attractive because of set of simple
806: resources can be used. Light
807: is usually presented by its modes, which are specified by its wave
808: vectors $\mathbf{k}$ or "which path", polarization, say
809: horizontally $H$ and vertically $V$, and occupation number. So
810: $\ket{2H}_{\mathbf{k}}$ is a
811: $H$-polarized mode with wave vector $\mathbf{k}$ and occupation number 2,
812: but for shortness they often say about two $H$ photons, that pass
813: along $\mathbf{k}$ direction or belong to the same space mode.
814: In proposals and experiments two
815: types of multimode states are discussed. First of them has one
816: photon distributed into $n$ modes
817: \begin{eqnarray}
818: \label{Wn1}
819: % \nonumber to remove numbering (before each equation)
820: W_{n}(1)=1/\sqrt{n}(\ket{1\dots 00}+\dots \ket{0\dots 01}),
821: \end{eqnarray}
822: and second known as polarized $W$ state reads
823: \begin{eqnarray}
824: \label{WnV}
825: W_{n}(V)=1/\sqrt{n}(\ket{V\dots HH}+\dots \ket{H\dots HV}).
826: \end{eqnarray}
827: The presented definitions are directly generalized to
828: $\eta_{n}(1)$ and $\eta_{n}(V)$.
829: \\
830: From point of the view of QIT, that considers logical qubits,
831: both states (\ref{Wn1}) and (\ref{WnV}) are equivalent up to
832: labelling $0\leftrightarrow H, 1\leftrightarrow V$, therefore they
833: have the same entropy, degree of entanglement and so on. Also
834: statistics of light in these states are similar. For $W_{n}(1)$
835: one finds the next correlation functions
836: \begin{eqnarray}
837: \nonumber
838: \mean{a_{k}}=\mean{a_{k}a_{m}}=0,&&\\
839: \mean{a^{\dagger}_{k}a_{m}}=
840: \mean{a^{\dagger}_{k}a_{k}a^{\dagger}_{m}a_{m}}=1/n.&&
841: \end{eqnarray}
842: It follows, that each of the modes has subpoissonian statistics
843: of photons with the Mandel parameter $\xi=-1/n$. In modes the
844: photons are anti-correlated because the coincident rate of the
845: photon counting $\mean{a^{\dagger}_{k}a_{k}a^{\dagger}_{m}a_{m}}$
846: is less then product
847: $\mean{a^{\dagger}_{k}a_{k}}\mean{a^{\dagger}_{m}a_{m}}$. All
848: these properties are true for $W_{n}(V)$.
849: \\
850: There is a difference between these states. Indeed all current
851: proposals based on the linear lossless optical elements permit
852: only conditional preparation of $W_{n}(V)$ in contrast to
853: $W_{n}(1)$.
854: %\\
855: %\underline{Observation?} Using a multiport device consisting from
856: %linear lossless and passive optical elements as BS, PBS etc, that
857: %is described by an unitary transformation, the polarized
858: %$W_{n}(V)$ states can only be prepared conditionally in contrast
859: %to $W_{n}(1)$.
860: %\\
861: \\
862: Considering the optical schemes proposed for generation of the $W$
863: states, one finds a similar structure of them. Their main
864: resources are linear optical elements $U$, sources of light $S$
865: and photon detectors $D$. A set of the linear elements is
866: presented mainly by beamsplitters ($BS$), polarized beamsplitters
867: ($PBS$), half- and quarter-wave plates (HWP, QWP), and others.
868: These devices are passive, conserve the number of photons and can
869: be described by an unitary transformation $U$. One of the most
870: popular sources of light
871: is the type-I and the type-II parametric down convertor ($PDC$)
872: in the threshold regime known as spontaneous parametric source
873: (SPDC), also the single-photon
874: source ($SPS$) is often discussed. For several experimental
875: proposals it requires commercial photon detectors, which can't
876: resolve the photon number of detection. So that these elements are
877: seem to be feasible by current technologies.
878: \\
879: Each scheme has two partitions at last. First splits the photons
880: of sources by linear optical elements into the output photons $O$
881: and the working photons $M$:
882: \begin{eqnarray}
883: % \nonumber to remove numbering (before each equation)
884: S\to U_{s} \to [O-M].
885: \end{eqnarray}
886: In the second partition after some unitary transformations all $M$
887: photons come to detectors and $O$ photons leave the schemes:
888: \begin{eqnarray}
889: % \nonumber to remove numbering (before each equation)
890: \leftarrow U_{O} \leftarrow [O-M] \to U_{M} \to D.
891: \end{eqnarray}
892: The key idea is simple. A given set of linear optical elements
893: entangles photons of the source and distributes them so that their
894: superposition contains a desired state of $O$ photons. It is
895: extracted with a probability by a projection measurement on $M$
896: photons. The probability of the successful outcome depends on the
897: resources used and is one of the main characteristic of these
898: schemes.
899:
900: \subsubsection{Experimental proposals}
901:
902: The above abstract arguments can be found by examining several
903: experimental schemes proposed.
904: \\
905: First note, that a scheme for $W$ state using third order
906: nonlinearity for path entangled photons has been introduced by
907: Zeilinger et al. \cite{ZeilingerNASA Conf}.
908: \\
909: In the scheme for the generation of $W_{4}(V)$ introduced by
910: Mathis et al. \cite{Mathis} the presented setup consists of the
911: type-II $PDC$ and two $SPS's$ as input modes. With the help of the
912: post-selection strategy developed for $GHZ$ \cite{Pan} the
913: $W_{4}(V)$ state can be achieved with probability 2/27 also
914: $W_{3}(V)$ can be done.
915: \\
916: An example of preparation of $W_{n}(1)$ with probability 1 and
917: $W_{3}(V),W_{4}(V)$ is given by Tomita et al. \cite{Bao-Sen Shi}.
918: The main resources are a set of $n$ single-photon sources and a
919: lossless $n\times n$ multiport fiber beamsplitter. If one photon
920: enters the input of the multiport beamsplitter the output state is
921: $\eta_{n}(1)$
922: or particular $W_{n}(1)$ for symmetric configuration
923: because of one photon is distributed with $n$ output
924: modes
925: \begin{eqnarray}
926: U_{n}:~~\ket{10\dots 0}\to W_{n}(1).
927: \end{eqnarray}
928: This method is generalized to $W_{n}(V)$ but only conditional
929: schemes can be achieved. One of the reasons of it is that already
930: two input photons are transformed by a beamsplitter into three
931: states $a\ket{11}+b\ket{20}+c\ket{02}$, however two of them $
932: \ket{20}, \ket{02}$ are often unwanted. Then it needs a
933: post-selection. Let three photons $H,H,$ and $V$ enter the input
934: of a tritter $U_{3}$ at the same time, then
935: \begin{eqnarray}
936: \label{Tritter}
937: U_{3}:~~
938: \ket{1H}_{1}\ket{1H}_{2}\ket{1V}_{3}\to aW_{3}(V)+\dots.
939: \end{eqnarray}
940: If we select outcome in which there is only one photon in each
941: output then the polarized $W_{3}(V)$ states are obtained with
942: probability 1/9. The found setup results in $W_{4}(V)$ with
943: probability 1/16, which is larger, than 2/27 of Ref. \cite{Mathis}
944: \\
945: Two schemes for $W_{3}(V)$ with type-II PDC and a set of SPS's is
946: discussed by Yamomoto et al. \cite{Yamamoto}. In the first scheme
947: beamsplitters transform the four photon states of PDC into a
948: superposition of the form
949: \begin{eqnarray}
950: \ket{2H}_{\mathbf{k}}\ket{2V}_{\mathbf{k}}\to \alpha
951: \ket{1V}_{\mathbf{k}} W_{3}(V)+\ket{1H}_{\mathbf{k}} W_{3}(H)
952: \end{eqnarray}
953: Then after a projection measurement on the working photon in
954: $\mathbf{k}$ mode the $W_{3}(V)$ or $W_{3}(H)$ is prepared with
955: maximal probability 3/32. The second scheme includes three SPS's
956: and is similar to (\ref{Tritter}).
957: \\
958: Using schemes proposed by Kobayashi et al. in ref. \cite{Kobayashi
959: multiphoton entangled states} the $W_{4}(V)$, GHZ and ZSA states
960: are generated. The experimental setup includes type-I and type-II
961: PDC. In the first scheme the state of sources is transformed by a
962: tritter
963: \begin{eqnarray}
964: (\ket{1H}_{a}\ket{1V}_{b}+\ket{1V}_{a}\ket{1H}_{b})
965: \ket{2H}_{c}\to aW_{4}(V)+\dots.
966: \end{eqnarray}
967: and projects onto a single photon state. By this way $W_{3}(V)$
968: can be done with probability 0.0165. In this scheme the type-I PDC
969: can be replaced by laser beam, from which the Fock state
970: $\ket{mH}$ originates. In the
971: second scheme the initial state involved two photons from each PDC
972: $\ket{4H}_{a}\ket{0}_{b}+\ket{2H}_{a}\ket{1H1V}_{b}+\ket{0}_{a}
973: \ket{2H2V}_{b}$ are transformed into $W_{4}(V)$ or to ZSA state.
974: Indeed, in these schemes all photons are working and are detected.
975: \\
976: Type-I PDC of a two-crystal geometry is proposed by Kobayashi for
977: generation of four-photon entanglement \cite{Kobayashi
978: two-crystal}.
979: \begin{figure}[ht]
980: \label{Bs}
981: \centering
982: \epsfxsize=8cm \epsfbox{4WKobayasha.eps}
983: \caption{Scheme for generation of $W_{4}(V)$
984: using four-photon states $\alpha\ket{4H}+\beta\ket{4V}+\gamma\ket{2H2V}$ of
985: Type-I PDC of a two-crystal geometry}
986: \end{figure}
987: The scheme presented at fig. 1. The source produces the
988: four-photon state $a\ket{4H}+b\ket{4V}+c\ket{2H2V}$. Experimental
989: setup includes beamsplitters, that transform the state of source
990: into entangled $a W_{4}(H)+b W_{4}(V)$ at a, b, c and d outputs.
991: To extract $W_{4}(V)$ there is a set of beamsplittes $BS_{v}$,
992: which have a small transparency for $V$ photons. By this way the
993: obtained state of the transmitted and reflected photons reads
994: $W_{4}(H)$ and $W_{4}(V)$ and is achieved by projective
995: measurement. Consider efficiency of these scheme. It can be
996: calculated assuming, that the probability of generation a photon
997: pair is about $\nu=4\times 10^{-4}$ per pulse. Probability of the
998: four-photon events the probability has order of $\nu^{2}=8\cdot
999: 10^{-8}$. For a pump laser with a 100-MHz repetition rate the
1000: generation rate of $W_{4}(V)$ is about $10^{-1}$ per second and
1001: with the 50\% detection efficiency one finds 1 desired state per
1002: minute.
1003:
1004: \subsection{Experiments}
1005:
1006:
1007: The recent experiments for generating multiphoton entanglement are
1008: based on PDC and linear optics elements manipulating polarized
1009: light.
1010: \\
1011: Weinfurter et al have proposed a type-II PDC
1012: \cite{WeinfurterZukowski} that has been used as a source in many
1013: experiments. Its state is a superposition of the four-photon $GHZ$
1014: and the tensor product of two maximally entangled EPR pairs,
1015: emitted into the two spatial modes
1016: \begin{eqnarray}
1017: \Psi^{(4)}=\sqrt{2/3}\ket{GHZ}-\sqrt{1/3}\ket{EPR}
1018: \ket{EPR}.
1019: \end{eqnarray}
1020: Here $\ket{EPR}=(\ket{H}_{a}\ket{V}_{b}+
1021: \ket{V}_{a}\ket{H}_{b})/\sqrt{2}$, and $GHZ$ has two
1022: indistinguishable photons of the same polarization into one space
1023: mode : $\ket{GHZ}=(\ket{2H}_{a}\ket{2H}_{b}+
1024: \ket{2V}_{a}\ket{2V}_{b})/\sqrt{2}$. This is not a product of two
1025: entangled pairs. If each of two modes splits at a beamsplitters,
1026: correlations between four photons can be observed \cite{Eibl
1027: four-photon}. However it needs to select events such that one
1028: photon is detected in each of the four modes. Two types of
1029: coincidence due from $GHZ$ and $EPR$ state between 300 and 100 per
1030: hour for integration time of 5 and 17.5 h. have been observed. By
1031: this way the Bell inequality for four qubits has been tested.
1032: \\
1033: Using the source generated $\Psi^{(4)}$, Weinfurter et al
1034: \cite{EiblThreeQubit} have generated the $W_{3}(V)$ state and
1035: examined its entanglement. The experimental setup is given at fig.
1036: 2.
1037: \begin{figure}[ht]
1038: \label{Bs}
1039: \centering
1040: \epsfxsize=8cm \epsfbox{3Wweinf.eps}
1041: \caption{Experimental setup to generate $W_{3}(V)$ by the SPDS source
1042: state of $\Psi^{(4)}=\sqrt{2/3}\ket{GHZ}-\sqrt{1/3}\ket{EPR}
1043: \ket{EPR}$ }
1044: \end{figure}
1045: Four photons in the $\Psi^{(4)}$ state are distributed into four
1046: modes $t,a,b,$ and $c$. By detection a photon in each of the arms
1047: the $W_{3}(V)$ entanglement in $a,b,c$ modes is prepared. In
1048: experiment several characteristics have been tested to verify the
1049: observed state. First is the coherence arisen from of the wave
1050: function. The reason is that the desired state is produced by
1051: independent measurements of the modes and coincidence
1052: photocurrents are examined. However by this way coherence of the
1053: wave function can't be detected and both the pure state and mixed
1054: one can't be distinguished. Second is the robustness of the $W$.
1055: In experiment it has been tested by performing an one-mode
1056: measurement, that projects a mode in to $H$ or $V$ state. It
1057: needs two measurements of the such type to destroy entanglement.
1058: So that after projecting in $\ket{H}_{a}$ an EPR pair of the form
1059: $1/\sqrt{2}(\ket{HV}+\ket{VH})_{bc}$ has been observed. Also the
1060: generalized Mermin inequalities \cite{Mermin} has been examined.
1061: However Cereceda has pointed out, that these inequalities can't
1062: verify the tripartite entanglement \cite{Cereceda}
1063: \\
1064: A problem of detection of genuine multiparticle entanglement of
1065: $W_{3}(V)$ and $\Psi^{(4)}$ has been studied experimentally by H.
1066: Weinfurter et al. \cite{Weifurter Witness}.
1067: A set of witness
1068: operators has been measured for $W$ and $\Psi^{(4)}$ using a set
1069: of polarization analyzers which consist of QWP, HWP and PBS. One
1070: of the witness for $W$ is given by (\ref{Ww1}), where the spin
1071: observable $\sigma_{z}$ corresponds to measurement of $H$, $V$
1072: linear polarization, $\sigma_{x},\sigma_{y}$ corresponds to
1073: analysis of $\pm 45^{0}$ linear polarization (left-right circular
1074: polarization). By this way the genuine $W$ entanglement has been
1075: demonstrated. Indeed for four -photon state, 15 different analyzer
1076: settings are required.
1077: \\
1078: Several quantum informational tasks can be done using the
1079: four-photon state $\Psi^{(4)}$. Weifurter et al. \cite{Weifurter
1080: DF} have demonstrated preparation of decoherence-free states which
1081: enable to encode a qubit in decoherence-free space. The scheme is
1082: presented at fig. 3.
1083: \begin{figure}[ht]
1084: \label{Bs}
1085: \centering
1086: \epsfxsize=10cm \epsfbox{DFweinf.eps}
1087: \caption{Experimental setup for decoherence-free communication.
1088: The photons of the source are distributed into four mode a,b,c,d.
1089: The noisy channel is presented by quarter- (QWP) and half wave
1090: plates (HWP). Polarization analyzers (PA) are employed for the registration
1091: of the decoherence-free states.}
1092: \end{figure}
1093: In experiment two DF states of the form
1094: (\ref{DF1}) and (\ref{DF2}) were generated from $\Psi^{(4)}$ and
1095: sent into noisy channel. The channel is presented by inserting QWP
1096: and HWP in each modes. The invariance of the encoded information
1097: has been observed by comparing the density matrix before and after
1098: the interaction with environment.
1099: \section{Applications}
1100: Some information tasks need entanglement so that there is a
1101: question whether MES of the W-class can be used. It has been found
1102: that some of these states are suitable for the problem of secrete
1103: sharing and key destribution \cite{Jaewoo JooSS}, teleportation
1104: \cite{Jaewoo JooTel} and dense coding also the distillation
1105: protocols \cite{Zhuo-Liang Cao Ming Yang1}, \cite{Zhuo-Liang Cao
1106: Ming Yang2} have been proposed.
1107: \subsection{Quantum key distribution and secrete sharing} Key
1108: distribution and secrete sharing are problems of classical
1109: cryptography and can be implemented using quantum resources,
1110: particularly $W$ state and projective measurements. Let three
1111: parties Alice, Bob and Claire share the $W$ state given by
1112: (\ref{W}) and perform randomly measurement of $\sigma_{x}$ and
1113: $\sigma_{z}$ on his own particle. The key idea of using $W$ state
1114: is that after projecting onto $\ket{0}_{A}$ Bob and Claire have in
1115: their hands entanglement $(\ket{01}+\ket{10})/\sqrt{2}$ and the
1116: subsequent outcomes of their measurements will be correlated. In
1117: contrast after projecting onto $\ket{1}_{A}$ they have a product
1118: $\ket{00}$ and the independent outcomes aries. This is a basis of
1119: quantum key distribution (QKD) because two correlated outcomes
1120: represent a key bit in the Bob and Claire hands. Quantum secrete
1121: sharing (QSS) is a QKD among $n$ persons in such a way that one's
1122: key message can be retrieved by the $k\leq n$ persons if they
1123: cooperate all together.
1124: \\
1125: In more detail. To describe the measurement of $\sigma_{x}$ and
1126: $\sigma_{z}$ introduce their eigenvectors
1127: \begin{eqnarray}
1128: % \nonumber to remove numbering (before each equation)
1129: \sigma_{x}\ket{x_{\pm}}=x_{\pm}\ket{x_{\pm}},~~~\sigma_{z}\ket{z_{\pm}}=z_{\pm}\ket{z_{\pm}},&&
1130: \end{eqnarray}
1131: where $\ket{x_{\pm}}=(\ket{0}\pm\ket{1})/\sqrt{2},
1132: \ket{z_{\pm}}=\ket{0},\ket{1}$, and $x_{\pm},z_{\pm}=\pm 1$ are
1133: eigenvalues or outcomes of the measurement. In the case of
1134: projecting, for example, into $\ket{z_{+}}$ outcome of the
1135: measurement is $z_{+}$. Using these notations $W$ state can be
1136: written in the form
1137: \begin{eqnarray}
1138: \nonumber
1139: W=(1/\sqrt{3})[\ket{z_{+}z_{-}z_{-}}+\ket{z_{-}z_{+}z_{-}}
1140: \ket{z_{-}z_{-}z_{+}}]&&\\
1141: =(1/\sqrt{3})\ket{z_{+}}[\ket{x_{+}x_{+}}-\ket{x_{-}x_{-}}]
1142: +(1/2\sqrt{3})\ket{z_{-}}[\ket{x_{+}x_{+}}+\ket{x_{+}x_{-}}+
1143: \ket{x_{-}x_{+}}+\ket{x_{-}x_{-}}].&&
1144: \end{eqnarray}
1145: When Alice has outcome $z_{+}$ then Bob and Claire obtain
1146: the same outcome $x_{+}$ or $x_{-}$ which is a key bit.
1147: The success probability in distributing a key bit is 2/3 .
1148: The case when Alice has outcome $z_{-}$ is useless and discarded.
1149: Note, the difference in correlation between two states
1150: $\Psi^{+}=(\ket{01}+\ket{10})/\sqrt{2}$ and $\ket{00}$ in the Bob
1151: and Claire hands. It is presented here by the number of outcomes,
1152: which are $x_{+}x_{+};~ x_{-}x_{-}$ for $\Psi^{+}$ and
1153: $x_{+}x_{+};~ x_{+}x_{-};~x_{-}x_{+};~ x_{-}x_{-}$ for $\ket{00}$.
1154: So any correlation may reduce the total number of outcomes. The
1155: protocol requires 12 qubits per a key bit at average. On the other
1156: hand the protocol E91 (based on Bell's theorem and EPR pairs as a
1157: quantum channel) has the overall success probability 2/9 and it
1158: requires 9 qubits per bit.
1159: \\
1160: In QSS Bob and Claire are expected to retrieve message from
1161: Alice in their cooperation. If Alice has outcome $z_{+}$ then Bob
1162: and Claire have opposite outcomes out of $z_{+}$ and $z_{-}$.
1163: Otherwise both have the same outcome $z_{+}$. When Bob and Claire
1164: cooperate they can collect their outcomes to correctly deduct
1165: the key bit of A. The overall success probability is 1/8 which is
1166: determined by the probability of the choosing measurements. Due
1167: to this argument 24 qubits are necessary to share a key bit.
1168: %On the other hand HBB99 with probability 1/2
1169: %needs 6 qubits per a key bit .
1170: It has been shown that these protocols are secure against simple
1171: individual attacks by an eavesdropper. These attacks are such that
1172: Eve performs an unitary operation on a composite system of her
1173: auxiliary qubit and one of the three qubits which are involved in
1174: a secure communication and she tries to extract some information
1175: by measuring her auxiliary qubit \cite{Gisin}.
1176:
1177:
1178: \subsection{Distillation of W}
1179: When an entangled state is transmitted its quantum correlations
1180: can be destroyed because of noise. To achieve a faithful
1181: transmission Bennet at all have proposed purification of the
1182: state using LOCC \cite{Bennett Purification}. It can be done using
1183: a set of the Pauli and CNOT operations. In \cite{Zhuo-Liang Cao
1184: Ming Yang2} a protocol for distillation of the form
1185: \begin{eqnarray}
1186: % \nonumber to remove numbering (before each equation)
1187: a\ket{100}+b\ket{010}+c\ket{001}\to (1/\sqrt{3})
1188: (\ket{100}+\ket{010}+\ket{001})
1189: \end{eqnarray}
1190: has been proposed. It includes two points: 1/ unitary
1191: transformations $U_{k}, k=1,2$ on $W$ and ancilla qubit
1192: \begin{eqnarray}
1193: U_{k}=
1194: \begin{pmatrix}
1195: 1 & 0 & 0 & 0 \\
1196: 0 & v_{k} & 0 & \sqrt{1-v^{2}_{k}} \\
1197: 0 & 0 &-1 & 0 \\
1198: 0 & \sqrt{1-v^{2}_{k}} & 0 & -v_{k} \\
1199: \end{pmatrix},
1200: \end{eqnarray}
1201: where $v_{1}=c/a$ and $v_{2}=c/b$, 2/ a measurement on ancilla.
1202: The successful probability is $3c^{2}$. This protocol can be
1203: simulated in cavity QED.
1204:
1205:
1206:
1207: \subsection{Teleportation and Dense coding using W-channel}
1208:
1209: Quantum teleportation, that allows transmitting an unknown
1210: states, is attractive for communications also for computing as
1211: primitive for quantum computations \cite{Gottesman}. The protocol
1212: has been demonstrated experimentally for teleportation of
1213: polarized photon \cite {D.Bouwmeester}, coherent state of light
1214: \cite{Furusawa} and atom \cite{AtomTeleport}. In the standard
1215: protocol an unknown qubit state
1216: \begin{eqnarray}
1217: \varphi=\alpha\ket{0}+\beta\ket{1}
1218: \end{eqnarray}
1219: can be transmitted using an $EPR$ pair as a quantum channel, 2
1220: bits of classical information gained in the Bell-state measurement
1221: and Pauli matrixes, which are retrieval operators.
1222: %Indeed Werner
1223: %has shown that the same resources are suitable for the dense
1224: %coding \footnote{R.F. Werner (2000), {\it All Teleportation and
1225: %Dense Coding Schemes}, quant/ph 0003070.}.
1226: \\
1227: Instead of EPR pair Karlsson et al. \cite{Karlsson} have
1228: considered the tree-particle $GHZ$ entanglement for sending
1229: unknown qubit to two receivers. It can be done probabilistically
1230: because of the non-cloning theorem, which forbids copying of an
1231: unknown states by linear unitary transformations \cite{Zurek}.
1232: However one of the receivers can retrieve unknown state if he'll
1233: cooperates with other receiver.
1234: \\
1235: Can the $W$ states be suitable for teleportation as a quantum
1236: channel? In a large number of the presented protocols the task of
1237: transmitting unknown qubit is accomplished probabilistically only
1238: \cite{ShiTomita}, \cite{Jaewoo JooTel}. Nevertheless there is an
1239: unconditional protocol for teleportation of entangled state
1240: \begin{eqnarray}
1241: \label{Ent}
1242: \phi=\alpha\ket{01}+\beta\ket{10}
1243: \end{eqnarray}
1244: which has been proposed in our Ref. \cite {GorbachevTrubilko}. It
1245: based on the observable, that two entangled qubits can be
1246: transmitted perfectly in the $GHZ$-channel by 3 bits of classical
1247: information if a Bell-like state measurement \cite{Luca} is
1248: performed
1249: \begin{eqnarray}\label{GHZTel}
1250: \phi_{12} \otimes |GHZ\rangle_{ABC}
1251: =(1/\sqrt{8})\sum_{x}|\Phi_{x}\rangle_{12A}[B_{x}\otimes
1252: C_{x}]\phi_{BC},%&&
1253: \end{eqnarray}
1254: where each of the eight vectors $\Phi_{x}$ is a product of the
1255: Bell state and eigenvector of $\sigma_{x}$, the retrieval
1256: operators $B$, $C$ are defined by Pauli matrixes. This equation
1257: tells, that if the GHZ channel allows teleportation of a state
1258: $\phi$, then this state can be teleported using any channel,
1259: obtained from the GHZ one by unitary transformation, that involves
1260: all particles of the channel except one.
1261: \\
1262: The required two-particle transformation reads
1263: \begin{eqnarray}\label{T203}
1264: V=|\Psi^{+}\rangle\langle 00|+ |11\rangle\langle 01| %&&\\\nonumber
1265: +|\Psi^{-}\rangle\langle 10|+|00\rangle\langle 11|.&&
1266: \end{eqnarray}
1267: It is a non-local unitary operation, that convert $GHZ$ into a
1268: state from the $W$ class
1269: \begin{eqnarray}\label{T2030} (1\otimes
1270: V)|GHZ\rangle_{ABC}=
1271: (1/\sqrt{2})\ket{100}+(1/2)\ket{010}+(1/2)\ket{001}.
1272: \end{eqnarray}
1273: By applying this transformation to both sides of (\ref{GHZTel}) we
1274: have teleportation of entangled state by the channel of the $W$
1275: class. There is new feature of the recovering operators which
1276: become non-local. The obtained $W$ channel can accomplish
1277: tree-qubit dense coding, when three bit of classical information
1278: can be send by manipulating two qubits \cite{3QubitC},
1279: \cite{GTRZ}.
1280: \section{Conclusions}
1281: Considering MES of the $W$-class we find interesting properties,
1282: proposals of their implementations and experimental realizations.
1283: They can be used for teleportation and dense coding also in
1284: quantum quantum cryptography for key distribution, secrete sharing
1285: and others. One of the important features of these states, that
1286: follows from their entanglement, is robustness, which
1287: distinguishes them from another states. So it has been shown that
1288: they have immunity to decoherence being decoherence-free states.
1289: However exploiting of this properties is not easy problem and we
1290: think that one of the main open questions is how to use fully the
1291: potential of MES from the $W$ class.
1292: \\
1293: \section*{Acknowledgments}
1294: We are grateful to A.M. Basharov for hopeful discussion. This work
1295: was supported in part by the Delzell Foundation Inc.
1296:
1297: \begin{thebibliography}{50}
1298: \bibitem{EntExp}
1299: K. Mattle, H. Weinfurter, P.G. Kwiat, and A. Zeilenger.
1300: Phys.Rev.Lett., {\bf 76}, 4656-4659 (1996).
1301: \\
1302: D.S. Naik, C.G. Peterson, A.G. White, A.J. Berglund, and P.G.
1303: Kwiat. Phys.Rev.Lett., {\bf 84}, 4733-4736 (2000).
1304: \\
1305: P.G. Kwiat, K. Mattle, H. Weinfurter, A. Zeilinger, A.V.
1306: Sergienko, Y.H. Shih. Phys.Rev.Lett., {\bf 75}, 4337-4341 (1995).
1307: \\
1308: P.G. Kwiat, E. Waks, A.G. White, I. Appelbaum, and P.H. Eberhard.
1309: Phys.Rev.A, {\bf{60}}, R773-R776 (1999).
1310: \\
1311: Y.H. Kim, M.V. Chekhova, S.P. Kulik, M. Rubin, Y.H. Shih. Phys.
1312: Rev. A {\textbf{63}}, 062301, (2001).
1313: \\
1314: G. Brida, M. V. Chekhova, M. Genovese, M. Gramegna, L. A.
1315: Krivitsky, and S. P. Kulik. Phys. Rev. A 70, 032332 (2004).
1316: \\
1317: Yu. I. Bogdanov, M. V. Chekhova, S. P. Kulik, G. A. Maslennikov,
1318: A. A. Zhukov., C. H. Oh, and M. K. Tey. Phys. Rev. Lett. 93,
1319: 230503 (2004).
1320: \bibitem{Cirac W} W. Dur, G. Vidal, J.I
1321: Cirac. Phys. Rev., A {bf{62}}, 062314, (2000)
1322: \bibitem{Polzik Nature}
1323: B. Julsgaard, A. Kozhekin, E.S. Polzik. Nature, {\bf{413}}, 400,
1324: (2001)
1325: \bibitem{Weinfurter Witness}
1326: M.Bourennane, M. Eibl, C. Kurtsiefer, H. Weinfurter, O. Guehne, P.
1327: Hyllus, D. Bruss, M. Lewenstein, A. Sanpera. Phys. Rev. Lett.,
1328: {\bf{92}}, 087902 (2004)
1329: \bibitem{IonExpW}
1330: C.F. Roos, M. Riebe, H. H\"{a}ffner, W.H\"{a}nsel, J.Benhelm, G.
1331: P.T. Lancaster, C. Becher, F.Schmidt-Kaler, R. Blatt. Scince,
1332: {\textbf{304}}, 1478, (2004).
1333: \bibitem{Dicke}
1334: R. Dicke. Phys. Rev, {\bf{93}}, 99 (1954)
1335: \bibitem{Bennett tel}
1336: C.H. Bennett, G. Brasard, C. Crepeau, R. Jozsa, A. Peres, W.K.
1337: Wootters. Phys. Rev. Lett., {\bf{70}}, 1895 (1993)
1338: \bibitem{Karlsson Rew 3W 3GHZ}
1339: A. Karlsson, M.Bourennane, I.Ghiu, D. Ljunggren, A. Manson.
1340: Quantum Computers and Computing, {\bf{4}}, 3 (2003)
1341: \bibitem{C.H. Bennett9511030}
1342: C.H. Bennett, H.J. Bernstein, S. Popescu, B. Shumacher. E-print,
1343: LANL, quant-ph/9511030 (1995).
1344: \bibitem{Greenberger Bell's theorem}
1345: D. M.Greenberger, M. Horne, A. Zeilinger. Bell's theorem, Quantum
1346: Theory and Conceptions of the Universe, ed. M. Kafatos, Kluwer, Dordrecht , 1989, p.69.\\
1347: D. Bouwmeester, J.-W. Pan, M. Daniell, H.Weinfurter, A.Zeilinger
1348: Phys. Rev. Lett., {\bf{82}}, 1345 (1999).
1349: \bibitem{Werner}
1350: R.F. Werner. Phys. Rev., A {\bf{40}}, 4277 (1989).
1351: \bibitem{Peres}
1352: A. Peres. Phys. Rev. Lett., {\bf{77}}, 1413 (1996).
1353: \bibitem{Horodecki Criterium}
1354: M. Horodecki, P. Horodecki, R. Horodecki. Phys. Lett. A,
1355: {\bf{223}}, 1 (1996).
1356: \bibitem{Pati}
1357: A.K. Pati. J. Physics, {\bf{59}}, 217 (2002).
1358: \bibitem{Zanardi Noiseless}
1359: P. Zanardi, M. Rasetti. Phys. Rev. Lett., {\bf{78}}, 3306 (1997).
1360: \bibitem{Askhat}
1361: %A.M.Basharov. \emph{Decoherence and entanglement in radiative
1362: %decay of a diatomic system.} JETP, {\bf{94}}, 1070 (2002)
1363: A.M. Basharov, JETP Lett 75, N3,123 (2002)\\
1364: A.M.Basharov and E.A.Manykin, Optics and Spectroscopy, 96, N1, 81
1365: (2004).
1366: \bibitem{Palma}
1367: G.M. Palma, K. Suominen, A.K. Ekert.
1368: Proc. R. Soc. London, Ser. A, {\bf{452}}, 567 (1996).\\
1369: L.-M. Duan, G.-C. Guo. Phys. Rev. Lett., {\bf{79}}, 1953 (1997).
1370: \bibitem{Lidar}
1371: D.A. Lidar, I.L. Chuang, K.B. Whaley.
1372: Phys. Rev. Lett., {\bf{81}}, 2594 (1998).\\
1373: D.A. Lidar, D. Bacon, J. Kempe, K. B. Whaley. Phys. Rev. A,
1374: {\bf{61}}, 052307 (2000)
1375: \bibitem{Kempe}
1376: J. Kempe, D. Bacon, D.A. Lidar, K. B. Whaley. Phys. Rev. A,
1377: {\bf{63}}, 042307 (2001).
1378: \bibitem{Weifurter DF}
1379: M. Bourennane, M. Eibl, S. Gaertner, C. Kurtsiefer, A. Cabello, H.
1380: Weifurter. Phys. Rev. Lett., {\bf{92}}, 107901 (2004).
1381: \bibitem{BriegelRaussendorf}
1382: H.J. Briegel, R.Raussendorf. Phys. Rev. Lett., {\bf{86}}, 910,
1383: (2001).
1384: \bibitem{Terhal}
1385: B.M. Terhal. Phys. Lett. A, {\bf{271}}, 319 (2000).\\
1386: M. Lewenstein, B. Kraus, J. I. Cirac, P. Horodecki.
1387: Phys. Rev. A, {\bf{62}}, 052310
1388: (2000).\\ D. Brub, J. I. Cirac, P. Horodecki, F. Hulpke, B. Kraus,
1389: M. Lewenstein, A.Sanpera. J. Mod. Opt., {\bf{49}}, 1399 (2002).
1390: \bibitem{Weifurter Witness}
1391: M.Bourennane, M. Eibl, C. Kurtsiefer, H. Weinfurter, O. Guehne, P.
1392: Hyllus, D. Bruss, M. Lewenstein, A. Sanpera. Phys. Rev. Lett.
1393: {\bf{92}}, 087902 (2004).
1394: \bibitem{Gorbachev parametric type interaction}
1395: V.N. Gorbachev, A.I. Zhiliba, A.A.Rodichkina, A.I.Trubilko. Phys.
1396: Lett. A, {\bf{323}}, 339 (2004).
1397: \bibitem{Agarwal}
1398: A.Biswas, G. S. Agarwal. J. Mod. Opt., {\bf{51}}, 1627 (2004).
1399: \bibitem{Zhuo-Liang Cao W state via cavity QED}
1400: Z.-L. Cao, M. Yang. Int. Journal of Quantum Information,
1401: {\bf{2}}, 231 (2004).
1402: \bibitem{Yun-Feng Xiao}
1403: Y.-F. Xiao, Z.-F. Han, J. Gao, G.-C. Guo. E-print, LANL,
1404: quant-ph/0412202 (2004). \\
1405: P. Xue, G.-C. Guo. Phys. Rev. A, {\bf{ 67}}, 034302 (2003).
1406: \bibitem{Gorbachev entangled W-states atom}
1407: V.N.Gorbachev, A.I. Trubilko, A.A. Rodichkina, A.I. Zhiliba. Phys.
1408: Lett. A, {\bf{310}}, 339, (2003).
1409: \bibitem{X. Wang}
1410: X. Wang. Phys. Rev. A, {\bf{ 64}}, 012313 (2001).
1411: \bibitem{ZeilingerNASA Conf}
1412: A. Zeilinger, M. A. Horne, D. M. Greenberger. NASA Conf. Publ. No.
1413: 3135 (National Aeronautics and Space Adminstration), Code NTT,
1414: Washington DC, 1992, p.73.
1415: \bibitem{Mathis}
1416: XuBo Zou, K. Pahlke, W. Mathis. E-print, LANL, quant-ph 0202090,
1417: (2002).
1418: \bibitem{Pan}
1419: J.-W. Pan, D. Bouwmeester, M. Danlell, H. Weinfurter , A.
1420: Zeilinger. Nature, {\bf{403}}, 515 (2000).
1421: \bibitem{Bao-Sen Shi}
1422: B.-S. Shi, A. Tomita. E-print, LANL, quant-ph 0208170 (2002).
1423: \bibitem{Yamamoto}
1424: T. Yamamoto, K. Tamaki, M. Koashi, N. Imoto. Phys. Rev. A,
1425: {\bf{66}}, 064301 (2002).
1426: \bibitem{Kobayashi multiphoton entangled states}
1427: H. Mikami, Y. Li, T. Kobayashi. Phys. Rev. A, {\bf{70}}, 052308
1428: (2004).
1429: \bibitem{Kobayashi two-crystal}
1430: Y. Li, T. Kobayashi. Phys. Rev. A, {\bf{69}}, R020302 (2004).
1431: \\
1432: Y. Li, T. Kobayashi. Phys. Rev. A, {\bf{70}}, 014301 (2004).
1433: \bibitem{WeinfurterZukowski}
1434: H. Weinfurter, M. Zukowski. Phys. Rev. A, {\bf{64}}, 010102
1435: (2001).
1436: \bibitem{Eibl four-photon}
1437: M.Eibl, S. Gaertner, M. Bourennane, Ch. Kurtsiefer, M. Zukowski,
1438: H. Weinfurter. E-print, LANL, quant-ph/0302042 (2003).
1439: \bibitem{EiblThreeQubit}
1440: M. Eibl, N. Kiesel, M. Bourennane, C. Kurtsiefer, H.Weinfurter.
1441: Phys. Rev. Lett., {\bf{92}}, 077901 (2004).
1442: \bibitem{Mermin}
1443: N. D. Mermin. Phys. Today, {\bf{43}}, 9 (1990).
1444: \bibitem{Cereceda}
1445: J.L. Cereceda. E-print, LANL, quant-ph/0402198 (2004).
1446: \bibitem{Jaewoo JooSS}
1447: J. Joo, J. Lee, J. Jang, Y.-J. Park. E-print, LANL,
1448: quant-ph/0204003 (2002).
1449: \bibitem{Jaewoo JooTel}
1450: J. Joo, Y.-J. Park, S. Oh, J. Kim. New Journal of Physics,
1451: {\bf{5}}, 136, (2003)\\
1452: Z.-L. Cao, W. Song. E-print, LANL, quant-ph/0401054 (2004).
1453: \bibitem{Zhuo-Liang Cao Ming Yang1}
1454: Z.-L. Cao, M. Yang. J. of Physics B: Atomic, Molecular and Optical
1455: Physics, {\bf{36}}, 4245, (2003).
1456: \bibitem{Zhuo-Liang Cao Ming Yang2}
1457: Z.-L. Cao, M. Yang. Physica A, {\bf{337}}, 141-148 (2004).
1458: \bibitem{Gisin}
1459: N. Gisin, G. Ribordy, W. Tittel, H. Zbinden. Rev. Mod. Phys.,
1460: {\bf{74}}, 145 (2002)
1461: \bibitem{Bennett Purification}
1462: C. H. Bennett, G. Brassard, S. Popescu, B. Schumacher, J. A.
1463: Smoin, W. K. Wootters. Phys. Rev. Lett., {\bf{76}}, 722 (1996).
1464: \bibitem{Gottesman}
1465: D. Gottesman, I.Chuang. Nature, {\bf{402}}, 390 (1999).
1466: \bibitem{D.Bouwmeester}
1467: D.Bouwmeester, J.W.Pan, M.Mattle, M.Eible, H.Weinfurter,
1468: A.Zeilinger. Nature, {\bf{390}}, 575 (1997).\\
1469: D. Boschi, S.Branca, F. de Martini, L.Hardy, S.Popescu. Phys.
1470: Rev. Lett., {\bf{80}}, 1121 (1998).\\
1471: Y.-H.Kim, S.P.Kulik, Y.Shih. Phys. Rev. Lett., {\bf{86}}, 1370
1472: (2001).
1473: \bibitem{Furusawa}
1474: A. Furusawa, J. L. Sorensen, S. L. Braunstein, C. A.Fuchs, H. J.
1475: Kimble, E. S. Polzik. Science, {\bf{282}}, 706 (1998).
1476: \bibitem{AtomTeleport}
1477: M. Riebe, H. Haffner, C.F. Roos, W. Hansel, J. Benhelm, G.P.T.
1478: Lancaster, T.W. Korber, C. Becher, F. Schmidt-Kaler, D.F.V. James,
1479: R. Blatt. Nature, {\bf{429}}, 734 (2004). \\ M.D.Barrett, J.
1480: Chiaverini, T.Schaetz, J.Britton, W.M. Itano, J.D.Jost, E.Knill,
1481: C. Langer, D. Leibfrider, R.Ozeri, D.J.Wineland.
1482: Nature, {\bf{429}}, 737 (2004).\\
1483: A.Danatan, N.Treps, A.Bramati, M.Pinard. Phys. Rev. Lett.,
1484: {\bf{94}}, 050502 (2005).
1485: \bibitem{Karlsson}
1486: A. Karlsson, M.Bourennane. Phys. Rev. A {\bf{58}}, 4394 (1998).
1487: \bibitem{Zurek}
1488: W.K.Wotters, W.S.Zurek. Nature, {\bf{299}}, 802 (1982).
1489: \bibitem{ShiTomita}
1490: B.-S. Shi, A.Tomita. Phys. Lett. A, {\bf{296}}, 161 (2002).\\
1491: J. Joo , Y.-J.Park. Phys. Lett. A, {\bf{300}}, 324 (2002).\\
1492: B.-S. Shi, A.Tomita. Phys. Lett. A, {\bf{300}}, 538-539 (2002).
1493: \bibitem{GorbachevTrubilko}
1494: V.N.Gorbachev, A.I.Trubilko. JETP, {\bf{91}}, 894 (2000).
1495: \bibitem{Luca}
1496: L.Marinatto, T.Weber. Found. Phys. Lett., {\bf{13}}, 119 (2000).\\
1497: V.N.Gorbachev, A.I.Zhiliba, A.I.Trubilko. J. Opt. B: Quantum
1498: Semiclass. Opt. {\bf{3}}, 825 (2001).
1499: \bibitem{3QubitC} J.L. Cereceda. E-print LANL, quant-ph/
1500: 0105096 (2001).
1501: \bibitem{GTRZ}
1502: V.N.Gorbachev, A.I.Trubilko, A.A.Rodichkina. Opt. Spectr.,
1503: {\bf{94}}, 706 (2001).
1504: \\
1505: V.N.Gorbachev, A.I.Trubilko, A.A.Rodichkina,
1506: A.I.Zhiliba. Phys. Lett. A, {\bf{314}}, 267 (2003).
1507: \end{thebibliography}
1508:
1509: \end{document}
1510: