1:
2: %\documentclass[prl,twocolumn,showpacs,amsmath,amssymb,cite]{revtex4}
3: %\documentstyle[prl,aps,epsf,cite]{revtex4}
4: %\documentclass[prl,aps,cite,twocolumn]{revtex4}
5: %\documentclass[prl,aps,cite,twocolumn,showpacs]{revtex4}
6: \documentclass[prl,aps,cite]{revtex4}
7: %\documentstyle[preprint,aps,epsf,cite]{revtex}
8: %\documentclass[prl,aps,epsf]{revtex}
9:
10: \newcommand{\newc}{\newcommand}
11: \newc{\beq} {\begin{equation}}
12: \newc{\eeq} {\end{equation}}
13: \newc{\beqa} {\begin{eqnarray}}
14: \newc{\eeqa} {\end{eqnarray}}
15: \newc{\bs} {\section}
16: \newc{\no} {\\ \nonumber}
17: \def\PL{{\em Phys. Lett.}}
18: \def\PLA{{ Phys. Lett.} {\bf A} }
19: \def\NPB{{\em Nucl. Phys.} {\bf B} }
20: %\def\PRL{{\em Phys. Rev. Lett. }}
21: %\def\PRB{{\em Phys. Rev.} {\bf B} }
22: %\def\PRA{{\em Phys. Rev.} {\bf A} }
23: \def\PRL{{ Phys. Rev. Lett. }}
24: \def\PRB{{ Phys. Rev.} {\bf B} }
25: \def\PRA{{ Phys. Rev.} {\bf A} }
26:
27: \def\PR{{\em Phys. Rev.} }
28: \def\PREP{{\em Phys. Rep.} }
29: \def\RMP{{\em Rev. Mod. Phys. } }
30: \usepackage{graphicx}
31: %\usepackage{epsfig}
32: %\usepackage{amsmath}
33:
34: \begin{document}
35: %\draft
36: %\twocolumn[\hsize\textwidth\columnwidth\hsize\csname
37: %@twocolumnfalse\endcsname
38: %\baselineskip 26pt
39: \title{Quantum key distribution using superposition of the vacuum and single photon states}
40: \author{Jae-Weon Lee and Jaewan Kim }
41: \address{
42: School of Computational Sciences, Korea Institute for Advanced
43: Study, 207-43 Cheongryangri-dong, Dongdaemun-gu Seoul 130-012,
44: Korea}
45: \author{ Yong Wook Cheong }
46: \address{Quantum Photonic Science Research Center, Hanyang University, Seoul
47: 133-791, Korea}
48: \author{ Hai-Woong Lee}
49: \address{Department of Physics, Korea Advanced Institute of Science and
50: Technology, Daejon 305-701, Korea }
51: \author{ Eok Kyun Lee }
52: \address{Department of Chemistry, School of Molecular Science (BK 21),
53: Korea Advanced Institute of Science and Technology, Daejon
54: 305-701, Korea.}
55:
56: \date{\today}
57: %\maketitle
58:
59: \begin{abstract}
60: {\normalsize B92-type and BB84-type quantum cryptography schemes
61: using superposed states of the vacuum and single particle states
62: which are robust against PNS attacks are studied. The number of
63: securely transferred classical bits per particle (not per qubit)
64: sent in these schemes is calculated and found to have upper
65: bounds.
66: Possible experimental realizations using the cavity QED or linear optics are suggested.
67: }
68: \end{abstract}
69: \pacs{PACS:03.67.2a, 03.67.Dd, 03.67.Lx}
70: \maketitle
71: %
72: %]
73: \bs{I. Introduction}
74:
75:
76: Recent progress in theories and
77: experiments\cite{PRL89,PRA62,knill,experiment,leeyh}
78: of generation and manipulation of single photons
79: allows one to think the quantum information processing
80: utilizing single particles feasible.
81: The first
82: commercial application of quantum information science at a single
83: qubit level might be the quantum key distribution
84: (QKD)\cite{gisin}. In the typical QKD scheme a sender (Alice)
85: shares a secret key with a receiver(Bob) by sending superposition
86: of photon polarization states. However, one can encode information
87: not only in particle states but also in the vacuum as shown in
88: some QKD schemes (mainly in double ray
89: schemes)\cite{B92,ardehali,qcomp}. In other words the vacuum can
90: play a role of an information carrier as particles do. Two of
91: authors had suggested the quantum teleportation and the Bell
92: inequality test using single-particle entanglement which was
93: verified experimentally later\cite{leekim}. In this direction we
94: proposed\cite{myepr} the Ekert-type \cite{Ekert} single ray
95: quantum cryptography scheme using the entangled states of the
96: vacuum and the single particle state\cite
97: {czachor,bjork,Tan,hardy,santos,mann,gerry,michler,leekim,giorgi}.
98: The main purpose of this work is to present a single ray B92-type
99: \cite{B92} and a BB84-type\cite{BB84} quantum cryptography schemes
100: using superposed states of the vacuum and single photon states.
101: In our schemes the detection of the superposition state is
102: possible with cavity QED devices or linear optics devices and
103: single photon detectors. Consider a quantum memory with the state
104: $|\phi\rangle=\alpha|0\rangle+\beta|1\rangle$ which is a
105: superposition of the vacuum $|0\rangle$ and a single photon state
106: $|1\rangle$ (optical qubit) which can be prepared by a photon
107: source using parametric down conversion\cite{hardy} or linear
108: optics\cite{lund} with the optical state truncation. By choosing
109: $|\beta|$ small enough we can make the expected number of photons
110: in $|\phi\rangle$ (i.e., $|\beta|^2$) arbitrary small. It implies
111: that we can encode classical bit information in the superposition
112: of the vacuum and single photon state with $|\beta|\ll 1$, which
113: is very faint light. Does this also mean that the legitimate
114: participants can share their secret key through a QKD protocol
115: with an arbitrary faint light source to hide quantum channel
116: itself from eavesdroppers? We show that the answer is $no$ at
117: least for straight forward generalizations of B92 and BB84-type
118: QKD with the vacuum-photon superposed state considered in this
119: paper, and there are upper bounds for classical bits shared
120: between parties per particle sent ($K$ defined below, not
121: counting the vacuum) in these schemes.
122:
123: This paper is organized as follows.
124: In Sec. II we present a B92-type quantum cryptography scheme using the superposition of
125: the vacuum and single particle states. We also calculate the number of classical bits
126: transferred per particle.
127: In Sec. III we extend the arguments to a BB84-type scheme.
128: In Sec. IV possible experimental realizations of our schemes using cavity QED
129: and linear optics are presented, and a security analysis is given.
130: Finally, in Sec. V we present a concluding discussion.
131: \begin{figure}[htbp]
132: \includegraphics[width=7cm,height=5cm]{fig1.eps}
133: \caption[Fig1]{\label{fig1}
134: Schematic of the cavity QED apparatus used in
135: the B92-type quantum cryptography scheme using superpositions of the vacuum and the single photon state.
136: See text for detailed explanations.
137: }
138: \end{figure}
139:
140:
141: \bs{II. B92-Type Scheme}
142: Fig. 1 shows our
143: B92-type quantum cryptography scheme using a cavity QED device.
144: As is well known the B92 protocol exploits the fact that arbitrary two non-orthogonal states
145: can not be distinguished perfectly.
146: Basically our scheme with a superposition of the vacuum and single photon
147: is just the same as the B92 scheme except for the state and the measuring device used.
148: For clarification we describe the scheme.
149:
150: (i) Alice sends sequences of states randomly chosen
151: between two non-orthogonal states $|\phi_0\rangle$ and
152: $|\phi_1\rangle$ representing logical 0 and 1, respectively; \beqa
153: |\phi_0\rangle&=&\alpha_0|0\rangle+\beta_0|1\rangle,\no
154: |\phi_1\rangle&=&\alpha_1|0\rangle+\beta_1|1\rangle, \eeqa
155: with normalization $|\alpha_i|^2+|\beta_i|^2=1~(i=0,1)$.\\
156: (ii) At a photon arrival time Bob measures a projection
157: operator randomly chosen between $P_0$ and $P_1$;
158: \beqa
159: \label{p}
160: P_0&\equiv &1-|\phi_1\rangle\langle\phi_1|,\no
161: P_1&\equiv &1-|\phi_0\rangle\langle\phi_0|.
162: \eeqa
163: (iii) After a series of measurements Bob publicly
164: announces to Alice in which instances he obtained a positive result.
165: This happens only when Alice sends $|\phi_0\rangle$
166: and Bob measures $P_0$ or Alice sends $|\phi_1\rangle$
167: and Bob measures $P_1$.
168: In other words, with probability $1/2$ the state sent by Alice and
169: the projection operator are correlated.
170: In these cases,
171: applying projection $P_0~(P_1)$ to $|\phi_0\rangle~(|\phi_1\rangle)$, Bob
172: obtains a positive result
173: with a probability\cite{POVM}
174: \beq
175: p=\frac{1}{2}(1-|\langle \phi_0|\phi_1\rangle|^2)=
176: \frac{1}{2}(1-|\alpha_0^* \alpha_1+\beta_0^* \beta_1|^2).
177: \eeq
178: Thus, after $N$ trials, the total $n_b\equiv pN < N$ bits of keys
179: are successfully shared, if there have been no eavesdropping or errors.
180:
181: (iv) To certify the absence of an eavesdropper Alice and Bob
182: sacrifice parts of data to check whether Bob obtained positive
183: results on $P_0~(P_1)$ measurement or not even in the case that
184: Alice sent $|\phi_1\rangle~(|\phi_0\rangle)$.
185:
186: At this point one can pose an interesting question.
187: How many classical bits can Alice
188: transfer per particle sent to Bob in an ideal case without errors or eavesdropping
189: up to the step (iii)?
190: According to the Holevo's
191: theorem\cite{holevo,capacity} asymptotically
192: one cannot encode
193: and retrieve reliably more than one bit of
194: classical information per $qubit$.
195: Note, however,
196: that in this letter we are interested in classical bit information not per qubit but per particle excluding
197: the vacuum, so
198: the bit information per $particle$ is not restricted by the Holevo bound.
199: In our schemes the number of qubits sent is not
200: equal to the number of particles sent, because the states sent are superpositions of the vacuum and
201: single particle states.
202: Let us calculate the ratio $K$.
203: Since Alice should choose randomly between
204: $|\phi_0\rangle$ and $|\phi_1\rangle$,
205: the density matrix for a transmission
206: can be written as $\rho=(|\phi_0\rangle \langle \phi_0|+
207: |\phi_1\rangle \langle \phi_1|)/2$.
208: So the average number of particles sent to Bob
209: is
210: \beq
211: n_p= N~Tr( \rho{\hat{n)}}=\frac{N}{2}(|\beta_0|^2+|\beta_1|^2) \le N
212: \eeq
213: where $\hat{n}$ is the particle number operator.
214: Then the ratio of bits transferred successfully to the average
215: number of particles
216: sent is therefore
217: \beq
218: K\equiv\frac{n_b}{n_p}=\frac{1-|\alpha_0^* \alpha_1+\beta_0^* \beta_1|^2}{
219: |\beta_0|^2+|\beta_1|^2}.
220: \eeq
221: Without loss of generality, using
222: the Bloch representation $(\alpha_i,\beta_i)=(cos(\theta_i),e^{i\psi_i}sin(\theta_i)),~(i=0,1)$
223: we can rewrite $K$ as
224: \beqa
225: \label{K}
226: K&=&\frac{1-|cos(\theta_0)cos(\theta_1)+sin(\theta_0)sin(\theta_1)
227: e^{i\psi}|^2}{sin^2(\theta_0)+sin^2(\theta_1)}\no
228: &\le&
229: \frac{1-\left (|cos(\theta_0)cos(\theta_1)|-|sin(\theta_0)sin(\theta_1)|\right )^2}{sin^2(\theta_0)+sin^2(\theta_1)}
230: \no
231: &\equiv& K_{max} (\theta_0,\theta_1)
232: ,
233: \eeqa
234: where $\psi\equiv\psi_1-\psi_0$ and the equality of the second line is satisfied when $\psi=0$ or $\pi$ and $
235: cos(\theta_0)cos(\theta_1)$ is opposite in sign to $sin(\theta_0)sin(\theta_1)e^{i\psi}$.
236: The upper bound on $K$ achieves value 2 asymptotically when $sin(\theta_0)=\pm sin(\theta_1) \rightarrow 0$
237: (but still nonzero, See Fig. 2).
238: Note that the optimal states are near the vacuum but not the vacuum states.
239: In the case that only one of $sin(\theta_0)$ and $sin(\theta_1)$ is $0$
240: (i.e., when one of $|\phi_0\rangle$ and $|\phi_1\rangle$ is the vacuum state.), $K$ becomes $1$.
241: $K=2$ means that in our QKD scheme Alice and Bob can share secure 2 classical bits of information
242: per single quantum particle transfer on the average in the ideal situation
243: of no error or eavesdropping. The physical reason for this bound is that
244: the more we send the vacuum (i.e. $|\beta_i|\simeq 0$), the smaller $n_p$ is, but then it is harder to
245: distinguish $|\phi_1\rangle$ from $|\phi_0\rangle$.
246: One can improve the probability of correct classification into
247: \beq p=1-|\alpha_0^* \alpha_1+\beta_0^* \beta_1| \eeq by optimal
248: positive operator valued measures (POVM)\cite{POVM,POVM2} using
249: ancilla qubits. In this case a similar argument leads to \beq
250: K=\frac{n_b}{n_p}=2\frac{1-|\alpha_0^* \alpha_1+\beta_0^*
251: \beta_1|}{ |\beta_0|^2+|\beta_1|^2}, \eeq which also has a
252: maximum value $2$ under the same condition stated below Eq.
253: (\ref{K}). (However, we will not consider a specific realization
254: of the POVM measurement
255: in this paper.)
256: On the other hand, for ordinary QKD
257: schemes using ordinary particle states
258: $K$ is usually smaller than $1$, because
259: there are always discarded data (i.e., $n_b < N$) to prevent Eve from distinguishing
260: Alice's states perfectly, while
261: to represent a qubit one or more particles are required(i.e., $n_p\ge N$).
262: For example the typical B92 scheme using two non-orthogonal photon polarization states has
263: $n_b=N/2$ (because the probability to get correct sifted keys is 1/2) and $n_p=N$, hence $K=1/2$.
264: In this sense, one can say that our B92-type scheme requires
265: a relatively smaller (four times smaller) number of particles to be transferred to send
266: a given classical bit information
267: than the typical B92 quantum cryptography
268: schemes.
269: \begin{figure}[htbp]
270: %\epsfig{file=fig1.eps,width=8cm,clip=}
271: \includegraphics[width=8cm,height=7cm]{fig2.eps}
272: \caption[Fig2]{\label{fig2}
273: Contour plot of the upper bound on the $K$ values ($K_{max}$) as a function of
274: $sin(\theta_0)$ and $sin(\theta_1)$.
275: Gray level represents the
276: value from black(0) to white(2).
277: }
278: \end{figure}
279: This ratio $K$ is similar to the key ratio per energy \cite{energy,cabello} but not exactly equal to that,
280: because in QKD schemes usually a sender and a receiver consume additional energy to share their references
281: and to communicate publicly.
282: $K$ is more likely a ``key rate per brightness" which measure how dark the quantum channel is for
283: a fixed information transmission rate.
284: If we have a QKD scheme with very big $K$, we can use the scheme to hide the quantum
285: channel itself from eavesdroppers, who will encounter a problem to find out where and when the
286: quantum channel opens.
287: This fact can be especially useful for QKD schemes on a moving
288: satellite or free-space systems. The bound for $K$ for our scheme
289: is non-trivial. If $K=2$ is just the Holevo bound divided by $1/2$
290: (vaguely guessed proportion of particles in the vacuum-single
291: photon superposed states), then we should also have maximal $K=2$
292: for the quantum memory or ordinary quantum channel but this is not
293: the case (maximal $K=\infty$ for these cases). Furthermore the
294: optimal $K$ value for our scheme even does not corresponds to the
295: case where
296: the average particle number in the state is 1/2.
297: Meanwhile the fact that one can store or send multi-bits
298: information per particle is not so surprising. In fact, it is
299: shown that infinite information can be transferred through
300: quantum channel at the cost of infinite entropy\cite{energy}.
301: However, what is interesting here is that although one can store
302: or send infinite information per particle in the quantum memory or
303: through a quantum channel using the vacuum-single particle
304: superposition, there is a upper bound for the key rate per
305: particle using QKD schemes
306: with the superposition ( at least for the QKD schemes considered in this paper).
307:
308:
309: \bs{III. BB84-Type Scheme}
310:
311: Compared to the B92 scheme, the BB84 scheme is known to be
312: more robust against the state discrimination attack\cite{usd}.
313: It is straightforward to extend our consideration to the BB84-type scheme\cite{BB84}
314: with superposed states of the vacuum and single particles.
315: As in the typical BB84 scheme,
316: Alice sends one of four states from two classes $\{|\phi_0\rangle,|\phi_1\rangle\}$, and $\{|\phi'_0\rangle$
317: $|\phi'_1\rangle\}$ to Bob where
318: $\langle \phi_0|\phi_1\rangle=0= \langle \phi'_0|\phi'_1\rangle$
319: and $|\phi_i\rangle~(i=0,1)$ are not orthogonal to
320: $|\phi'_i\rangle$. Then, Bob measures one of four projection operators
321: $P_i=|\phi_i\rangle \langle \phi_i|$ or $P'_i=|\phi'_i\rangle \langle \phi'_i|$.
322: After basis reconciliation with Alice via a public channel Bob would get the classical bit
323: information with probability 1/2, hence $n_b=N/2$. The density matrix of Alice's particle
324: is
325: $\rho=\left(|\phi_0\rangle \langle \phi_0|+
326: |\phi_1\rangle \langle \phi_1| +|\phi'_0\rangle \langle \phi'_0|+
327: |\phi'_1\rangle \langle \phi'_1|\right)/4$.
328: Therefore, the ratio of bits shared to the number of particles sent in this scheme is
329: \beq
330: K=\frac{4}{2(|\beta_0|^2+|\beta_1|^2+|\beta'_0|^2+|\beta_1'|^2)},
331: \eeq
332: where $|\phi_i\rangle=\alpha_i|0\rangle+\beta_i|1\rangle$
333: and $|\phi'_i\rangle=\alpha'_i|0\rangle+\beta'_i|1\rangle$.
334: The orthogonality condition $\langle \phi_0|\phi_1\rangle=0=\langle \phi'_0|\phi'_1\rangle$ implies
335: $|\beta_0|^2+|\beta_1|^2=1=|\beta'_0|^2+|\beta'_1|^2$, so $K=1$ which is twice the value of
336: $K$ for BB84 schemes with ordinary particles, because for the ordinary BB84 protocol
337: $n_b=N/2$ and $n_p=N$.
338:
339:
340: \bs{IV. Apparatus and security}
341:
342: We may now proceed to the description of
343: the apparatuses for our schemes shown in Fig. 1 for the B92-type scheme and in Fig. 3 for the BB84-type
344: scheme.
345: The setups consist of Alice's photon source(S) for generation of
346: the superposition of the vacuum and single particle states, and
347: Bob's projective measurement device using either cavity QED (Fig.
348: 1) or linear optics (Fig. 3) which are considered by many
349: authors\cite{davidovich,moussa,freyberger}. In principle the
350: cavity QED devices and the linear optics devices can be used both
351: for the B92 or the BB84 scheme. The detectors are essentially the
352: same detectors we considered in our previous work\cite{myepr}, so
353: we will just briefly review here. Let us first consider Fig. 1. By
354: utilizing the parametric down conversion or coherent light, the
355: source(S) generates $\phi_0$ or $\phi_1$ on Alice's demand.
356: Assuming that at time $t=0$ a ground state atom
357: $|g\rangle$
358: is injected into the cavity $C$,
359: the total cavity-atom state is then
360: $|\psi(0)\rangle=|\phi_i\rangle|g\rangle$.
361: The interaction between atoms and photons
362: in the cavity $C$ are described by
363: the Jaynes-Cummings Hamiltonian.
364: With this Hamiltonian and by choosing interaction time appropriately
365: one can transfer the information of photon
366: states $|\phi_i\rangle$
367: to that of the atoms (See the references for details). Then the projective measurement on the photon
368: state $\alpha|0\rangle+\beta|1\rangle$
369: can be possible by adjusting appropriately
370: the field in the Ramsey zones ($R$)
371: such that the state
372: undergoes a unitary evolution to the state
373: which registers a click in the state-selective
374: ionization detector $D$\cite{davidovich,gerry}. So this setup
375: performs ultimately deterministic (i.e., with probability 1 for ideal cases)
376: projection on $|B\rangle\equiv\alpha|0\rangle+\beta|1\rangle$.
377: Let us find $|B\rangle$ such that $P_0$ in Eq. (\ref{p})
378: can be written as
379: $|B\rangle \langle B|$.
380: It should satisfies $\langle B| \phi_1 \rangle =0$,
381: because $P_0|\phi_1\rangle=0$.
382: In other words, to measure $P_0$
383: Bob should set the fields in the Ramsey zone
384: so that the input photon state with
385: $\alpha=\beta^*_1$ and $\beta=-\alpha^*_1$
386: (i.e. orthogonal to $|\phi_1\rangle$)
387: corresponds to the click on detector $D$.
388: Similarly Bob can measure $P_1$ by performing
389: projection on $\beta^*_0 |0\rangle -\alpha^*_0|1\rangle$.
390: \\
391:
392: \begin{figure}[htbp]
393: %\epsfig{file=fig1.eps,width=8cm,clip=}
394: \includegraphics[width=8cm,height=5cm]{fig3.eps}
395: \caption[Fig3]{\label{fig3}
396: Schematic of the linear optics apparatus used in
397: the BB84-type quantum cryptography scheme using the superpositions of the vacuum and
398: the single particle state.
399: See text for detailed explanations.
400: }
401: \end{figure}
402: On the other hand, the projective measurement for the BB84-type scheme using linear optics shown in Fig. \ref{fig3} is
403: non-deterministic in a sense that the measurement succeeds only probabilistically.
404: This setup is a modification of
405: the setup proposed in ref. \cite{lund}.
406: The beam splitter $BS$
407: performs the mode transformation
408: \beq
409: \label{trans}
410: \left (
411: \begin{array}{cc}
412: a'\\
413: b'
414: \end{array}
415: \right )=\left (
416: \begin{array}{cc}
417: \sqrt{R} & \sqrt{1-R} \\
418: -\sqrt{1-R} & \sqrt{R}
419: \end{array}
420: \right )
421: \left (
422: \begin{array}{cc}
423: a\\
424: b
425: \end{array}
426: \right ),
427: %\]
428: \eeq
429: where $R$ is the reflectivity of the beam splitter.
430: In second quantized notation, the general input state
431: shown in Fig. \ref{fig3} can be
432: written as
433: \beq
434: \label{bs1}
435: \psi=(\gamma+ \delta~a^{\dagger})
436: (\alpha+\beta~b^{\dagger})|0\rangle
437: \eeq
438: with normalization $|\alpha|^2+|\beta|^2=1$.
439: Here, $(\gamma+\delta~ a^{\dagger})|0\rangle$ is a known probe state, while
440: $\alpha|0\rangle+\beta|1\rangle$ is an unknown input state to be
441: measured.
442: By replacing $a$ and $b$ in
443: Eq. (\ref{bs1}) with $a'$ and $b'$ using Eq. (\ref{trans}),
444: we obtain the output state
445: %\beq
446: $
447: \psi=[
448: \alpha\gamma + \frac{\beta\delta}{2}
449: (a'^{\dagger2}- b'^{\dagger2})
450: + \sqrt{2}\beta\gamma a'^{\dagger}
451: ]|0\rangle,
452: $
453: %\eeq
454: if we set $R=1/2$ and $\alpha\delta= \beta \gamma$. Therefore,
455: by noting that the detector $D_a$ detects single photon and $D_b$
456: detects none, Bob can perform projective measurement on the
457: superposition state $\alpha|0\rangle +\beta|1\rangle$ with the
458: probability of success $2|\beta \gamma|^2\le 1/2$\cite{myepr}. To
459: detect $|1\rangle$ or $|0\rangle$ state we simply replace the beam
460: splitter and check whether the detector $D_a$ fires or not. Then,
461: one of four states
462: $\{|\phi_0\rangle,|\phi_1\rangle,|\phi'_0\rangle$
463: $|\phi'_1\rangle\}$ of the section III sent to Bob by Alice can be measured with this apparatus
464: for the BB84-type scheme.
465:
466: Let us now discuss the security of our schemes.
467: Basically our schemes follow the ordinary B92 and the BB84 schemes except for
468: the states and measuring devices used, so one can
469: simply adopt the well known security proof
470: for these ordinary schemes\cite{security} for our schemes also.
471: Another merit of our schemes is that since the superposition of the vacuum and one photon is
472: not a photon number eigenstate, our schemes are robust against the photon number
473: splitting (PNS) attacks\cite{pns}. (The PNS attacks restrict key rates and distance for many
474: practical QKD schemes such as typical B92 or BB84 QKD schemes with weak coherent states.)
475: Because, even in the case the eavesdropper (Eve) has multiple
476: copies of the state ($|\phi\rangle^{\otimes n}$) due to
477: imperfections of the light sources,
478: to do the PNS attack Eve should perform the photon number non-demolition measurement,
479: but our schemes use the superposed states of the vacuum and single
480: photon which is inevitably destroyed by any photon number
481: measurement. This allows Alice and Bob to detect Eve attempting
482: the PNS attack by publicly comparing parts of the qubits sent with
483: the qubits received. So our schemes present yet another way for
484: security against the PNS attacks different from the recently
485: proposed schemes\cite{newschemes}.
486:
487: For our schemes sending the pure vacuum ($|\phi\rangle=|0\rangle$) as a qubit has an intrinsic problem that
488: Bob can not distinguish the vacuum from channel loss.
489: But fortunately as described above the pure vacuum state is not the optimal
490: state for the maximal $K$ value. So there is no reason
491: to use the pure vacuum state as a qubit for our schemes and we
492: can avoid this problem by simply not using the pure vacuum state.
493: In a practical sense, it is experimentally interesting but
494: challenging to implement the detection of a superposition of the
495: vacuum and single-photon states\cite{lund,singledetect}. Recently,
496: there are many related experimental and theoretical
497: works about transferring quantum states using the cavity\cite{cavityexp}.
498:
499: \bs{V. Discussion}
500: In summary, we have proposed the B92 and the BB84-type
501: quantum key distribution schemes using superposed states of the vacuum and the single particle state
502: robust against PNS attacks.
503: We showed that in our QKD schemes using the vacuum-photon superposition states
504: the information transferred per
505: particle sent is bounded. So far it is unclear that this restriction has more profound physical reasons.
506: Therefore proving or disproving the existence of an exotic QKD protocol which
507: has a big value of $K$, that is, a QKD scheme with very faint light might
508: be an interesting subject.
509: \\
510: \\
511: \indent J. Lee was supported by part by the Korea Ministry of Science and Technology.
512: J. Kim
513: was supported by the Korea Research Foundation (Grant
514: No. KRF-2002-070-C00029).
515: H. W. Lee was supported by the Ministry of Science and Technology of Korea.
516: E. Lee
517: was supported by the Korea Research Foundation (
518: Grant No. KRF-GH16110).
519:
520: %\newpage
521: \begin{thebibliography}{99}
522: \bibitem{PRL89}
523: G. J. Milburn, Phys. Rev. Lett. 62 (1989) 2124.
524: \bibitem{PRA62}
525: J.C. Howell and J.A. Yeazell, Phy. Rev. A 62 (2000) 012102.
526: \bibitem{knill}
527: E. Knill, R. Laflamme and G. J. Milburn, Nature, 409 (2001)46
528: \bibitem{experiment}
529: C. K. Hong and L. Mandel, Phys. Rev. Lett. 56 (1986) 58;
530: J. Kim, O. Benson, H. Kan and Y. Yamamoto, Nature 397 (1999) 500;
531: C. Brunel, B. Lounis, P. Tamarat and M. Orrit,
532: Phys. Rev. Lett. 83 (1999) 2722.
533: \bibitem{leeyh}
534: H.G. Park, et al.,Science, 305 (2004) 1444.
535: \bibitem{gisin}
536: See, for example,
537: N. Gisin, G. Ribordy, W. Tittel, and H. Zbinden
538: Rev. Mod. Phys., 74 (2002) 145.
539: \bibitem{B92}
540: C. H. Bennett, Phys. Rev. Lett. 68 (1992) 3121.
541: \bibitem{ardehali}
542: M. Ardehali, \PLA 217 (1996) 301.
543: \bibitem{qcomp}
544: Xu-Bo Zou, K. Pahlke and W. Mathis,\PLA 311 (2003) 271.
545: \bibitem{leekim}
546: H.W. Lee and J.K. Kim
547: Phys. Rev. A 63 (2001) 012305;
548: E. Lombardi, F. Sciarrino, S. Popescu, and F. De Martini, \PRL 88 (2002) 070402.
549: \bibitem{myepr}
550: J.W. Lee, E.K. Lee, Y.W. Chung, H.W. Lee and J.K. Kim, \PRA 68 (2003) 012324.
551: \bibitem{Ekert}
552: A.K. Ekert, Phy. Rev. Lett. 67 (1991) 661.
553: \bibitem{czachor}
554: M. Czachor, \PRA 49 (1994) 2231;
555: D. Home and G.S. Agarwal, \PLA 209 (1995) 1.
556: \bibitem{bjork}
557: G. Bj\"{o}rk and P. Jonsson and L.L. S\'{a}nchez-Soto,
558: \PRA 64 (2001) 042106.
559: \bibitem{Tan}
560: S. M. Tan, D. F. Walls and M. J. Collet, Phys. Rev. Lett.
561: 66 (1991) 252;
562: S. M. Tan, M. J. Holland and D. F. Walls, Opt. Commun. 77 (1990)
563: 285.
564: \bibitem{hardy} L. Hardy, Phys. Rev. Lett. 73 (1994) 2279;
565: L. Vaidman, Phys. Rev. Lett. 75 (1995) 2063;
566: D. M. Greenberger, M. A. Horne and A. Zeilinger,
567: Phys. Rev. Lett. 75 (1995) 2064;
568: L. Hardy, Phys. Rev. Lett. 75 (1995) 2065.
569: \bibitem{santos}
570: E. Santos, Phys. Rev. Lett. 68 (1992) 894;
571: S. M. Tan, D. F. Walls and M. J. Collet, Phys. Rev. Lett.
572: 68 (1992) 895.
573: \bibitem{mann}
574: M. Revzen and A. Mann, Found. Phys. 26 (1996) 847.
575: \bibitem{gerry}
576: C. C. Gerry, Phys. Rev. A 53 (1996) 4583.
577: \bibitem{michler}
578: M. Michler, H. Weinfurter, and M. Zukowski, \PRL, 84 (2000) 5457.
579: \bibitem{giorgi}
580: G. L. Giorgi, quant-ph/0504150.
581: \bibitem{BB84}
582: C.H. Bennett and G. Brassard, in Proceedings of the IEEE International Conference on Computers,
583: Systems and Signal Processing, Bangalore, India (IEEE, New York, 1984) (1996).
584: \bibitem{lund}
585: M. Paris \PRA 62 (200) 033813;
586: A. P. Lund and T. C. Ralph \PRA 66, 032307 (2002).
587: \bibitem{holevo}
588: A. S. Holevo, Probl. Inf.
589: Transm. 9 (1973) 177;IEEE Trans. Inf. Theory 44 (1998) 269.
590: \bibitem{capacity}
591: D. W. Berry, Phys. Rev. A 71 (2005) 032334.
592: \bibitem{POVM}
593: I. D. Ivanovic, Phys. Lett. A 123 (1987) 257;
594: D. Dieks, Phys. Lett. A 126 (1988) 303;
595: A. Peres, Phys. Lett. A 128 (1988) 19;
596: A. Chefles, Phys. Lett. A 239 (1998) 339;
597: A. Chefles, Contemp. Phys. 41 (2000) 401.
598: \bibitem{POVM2}
599: C. W. Helstrom, {\it Quantum Detection and Estimation Theory} (Academic Press, New York, 1976);
600: M. Mohseni,A. M. Steinberg,and J. A. Bergou,Phys. Rev. Lett. 93
601: (2004) 200403.
602: \bibitem{energy}
603: C.M. Caves and P.D. Drummond, Rev. Mod. Phys. 66 (1994) 481;
604: H.P. Yuen and M. Ozawa, Phys. Rev. Lett. 70 (1993) 363.
605: \bibitem{cabello} For a related key rate definition see also
606: Ad$\acute{a}$n Cabello, Phys. Rev. Lett. 85 (2000) 5635.
607: \bibitem{usd}
608: B. Huttner, J. D. Gautier, et al., Phys. Rev. A 54 (1996) 3783.
609: \bibitem{davidovich}
610: L.Davidovich et al.,\PRA 50 (1994) R895.
611: \bibitem{moussa}
612: M.H.Y. Moussa and B. Baseia, \PLA 245 (1998) 335.
613: \bibitem{freyberger}
614: M. Freyberger, \PRA 51 (1995) 3347.
615: \bibitem{security}
616: D. Mayers, {\it Advances in
617: Cryptology. Proceedings of Crypto กฏ96} (Springer-Verlag,
618: New York, 1996) (1996) 343;
619: H.-K. Lo and H. F. Chau, Science 283 (1999) 2050;
620: P. Shor and J. Preskill, \PRL 85 (2000) 441;
621: K. Tamaki and N. L\"{u}tkenhaus, \PRA 69 (2004) 032316.
622: \bibitem{pns}
623: B. Huttner, N. Imoto, N. Gisin, and T. Mor, Phys. Rev. A
624: 51, 1863 (1995); G. Brassard, N. L\"{u}tkenhaus, T. Mor, and
625: B. C. Sanders, Phys. Rev. Lett. 85, 1330 (2000).
626: \bibitem{newschemes}
627: W. Hwang, \PRL 91 (2003) 057901;
628: H. Lo et al., \prl 94 (2005) 230504;
629: X. Wang et al., \prl 94 (2005) 230503;
630: V. Scarani et al., \PRL 92 (2004) 057901;
631: M. Koashi, \prl 93 (2004) 120501.
632: \bibitem{singledetect}
633: D. Pegg, L. Phillips, and S. Barnett, \PRL 81 (1998) 1604;
634: S. A. Babichev, J. Ries and A. I. Lvovsky.
635: Europhys. Lett., 64 (2003) 1;
636: M. Koniorczyk, Z. Kurucz, A. Gabris, and J. Janszky \PRA 62 (2000) 013802.
637: \bibitem{cavityexp}
638: A. Rauschenbeutel et al.,\PRA,64 (2001) R050301;
639: Asoka Biswas and G. S. Agarwal,\PRA 70 (2004) 022323.
640: \end{thebibliography}
641: \end{document}
642: