1: \documentclass[12pt]{iopart}

2: \usepackage{epsf}

3: \begin{document}

4: \title{Secure quantum channels with correlated twin laser beams}

5: \author{Constantin V. Usenko\dag\ and Vladyslav C. Usenko\ddag

6: \footnote[3]{To whom correspondence should be addressed (usenko@univ.kiev.ua)} }

7: 

8: \address{\dag\ National Shevchenko University of Kyiv, Department of Theoretical Physics, Kyiv, Ukraine}

9: 

10: \address{\ddag\ Institute of Physics of National Academy of Science, Kyiv, Ukraine}

11: 

12: \begin{abstract}

13: This work is the development and analysis of the recently proposed quantum cryptographic protocol,

14: based on the use of the two-mode coherently correlated states. The protocol is supplied with the

15: cryptographic control procedures. The quantum noise influence on the channel error properties is examined.

16: State detection features are proposed.

17: \end{abstract}

18: \pacs{03.67.Dd, 03.67.Hk, 42.50.Ar,42.50.Dv}

19: 

20: \section{Introduction.}

21: 

22: 

23: The goals of quantum cryptography and secure quantum communications 

24: \cite{qc1, qc2, qc3} can be achieved using various protocols, which were 

25: developed and realized \cite{entprot, fourexp} in the past years on the basis 

26: of the quantum entanglement \cite{ent1, entprot} 

27: of weak beams and the single \cite{single1, single2} or few photon states \cite{four}, 

28: mostly by means of adjusting and detecting their polarization angles \cite{polar}. 

29: 

30: Another method, based on the usage of the two-mode coherently correlated (TMCC) beams

31: was proposed recently \cite{tmcc}. In this case the secure cryptographic key is generated

32: by the laser shot noise and duplicated through the quantum channel. Unlike the single or few photon schemes, 

33: which require large numbers of transmission reiterations to obtain the statistically significant 

34: results, the TMCC beam can be intensive enough to make each single measurement 

35: statistically significant and thus to use single impulse for each 

36: piece of information, and remain cryptographically steady.

37: 

38: In this work we analyse the error properties of the secure quantum channels, based on the

39: TMCC-beams and propose some additions to the TMCC-based cryptographic protocol.

40: 

41: The two-mode coherently correlated state is the way we refer to the 

42: generalized coherent state in the meaning by Perelomov \cite{per}. Such state can 

43: be described by its presentation through series by Fock states:

44: 

45: \begin{equation}

46: \label{eq:tmcc}

47: \left| \lambda \right\rangle =\frac{1}{\sqrt{I_0\left(2\left|\lambda\right|\right)}}\sum_{n=0}^\infty {\frac{\lambda ^n}{n!}\left| {nn} 

48: \right\rangle } 

49: \end{equation}

50: 

51: Here we use the designation $\left| {nn} \right\rangle = \left| n 

52: \right\rangle _1 \otimes \left| n \right\rangle _2 $, where $\left| n 

53: \right\rangle _1 $ and $\left| n \right\rangle _2 $stand for the states of 

54: the $1^{st}$ and $2^{nd}$ modes accordingly, represented by their photon 

55: numbers. The states (\ref{eq:tmcc}) are not the eigenstates for each of the operators 

56: separately, but are the eigenstates for the product of annihilation 

57: operators: 

58: \begin{equation}

59: 	a_1 a_2 \left| \lambda \right\rangle = \lambda \left| \lambda 

60: 	\right\rangle . 

61: \end{equation}

62: Such states can also be obtained from the zero state:

63: 

64: \begin{equation}

65: \label{eq:tmccground}

66: \left| \lambda \right\rangle = \frac{1}{\sqrt{I_0\left(2\left|\lambda\right|\right)}}I_0 (\lambda a_1^ + a_2^ + )\left| 0 

67: \right\rangle 

68: \end{equation}

69: 

70: 

71: In this work we assume that two laser beams, which are 

72: propagating independently from each other, correspond to the two modes of 

73: the TMCC state. States of beams are mutually correlated (surely, the TMCC 

74: state can also be represented in another way, for example, as a beam 

75: consisting of two correlated polarizations).

76: 

77: 

78: 

79: \section{Beam measurement}

80: 

81: 

82: Let's examine any of the two TMCC beams separately. 

83: The intensity of the beam's radiation, registered by an observer is proportional to the 

84: mean of the $N = a^ + a$ operator, which is the number of the photons 

85: in the corresponding mode. 

86: The mean observable values, which characterize the results of the 

87: measurements of the beam are:

88: 

89: \begin{equation}

90: \label{eq:meann}

91: \left\langle {N } \right\rangle = \left\langle \lambda \right|a^ 

92: + a \left| \lambda \right\rangle , \left\langle {N^2 } \right\rangle = 

93: \left\langle \lambda \right|a^ + a a^ + a \left| 

94: \lambda \right\rangle 

95: \end{equation}

96: 

97: 

98: These characteristics are squared in field, and thus their mean values don't turn to 

99: zero (this fact is not specific for the TMCC-states, because the usual non-correlated 

100: states and processes, like the heat propagation, show the same properties).

101: 

102: Assuming the state expression (\ref{eq:tmcc}) we obtain

103: 

104: \begin{equation}

105: \label{eq:meann2}

106: \langle {N } \rangle = \frac{1}

107: {I_0 ( 2| \lambda | )}\sum_{n=0}^\infty n \frac{| \lambda |^{2n}}{n!^2} , \langle {N^2 } \rangle = \frac{1}

108: {I_0 ( 2| \lambda | )}\sum_{n=0}^\infty n^2 \frac{| \lambda |^{2n}}{n!^2}

109: \end{equation}

110: 

111: 

112: The mean number of registered photons is

113: 

114: \begin{equation}

115: \langle N \rangle = \sum_{n=0}^\infty nP_n (\lambda)

116: \end{equation}

117: 

118: 

119: 

120: The probability of registering n photons depends on the intensity of a beam:

121: 

122: \begin{equation}

123: \label{eq:nphotprob}

124: P_n(\lambda) = \frac{1}{I_0 (2|\lambda|)}\frac{|\lambda|^{2n}}{n!^2} 

125: \end{equation}

126: 

127: An important feature of this distribution is the quick (proportional to $n!^2$) decreasing dependence of the registration probability

128: on the photon number. 

129: 

130: This circumstance makes the experimental identification of the

131: TMCC-states quite convenient. The distribution of the probability of different photon 

132: numbers registration along with the analogous distribution for a usual coherent beam 

133: are given at the \fref{plot_pn}. One can see that there are significant differences 

134: for the TMCC and the Poisson beam distributions - the TMCC-beam distribution is relatively 

135: sharp and narrow.

136: 

137: 

138: 

139: \begin{figure}[htbp]

140: \begin{center}

141: 	\epsfbox{plot_pn.eps}

142: 	\caption{ Probability of different photon numbers registration distribution for the TMCC-beam (circles, solid line) and the analogous distribution for a usual coherent beam (squares, dotted line)}

143: 	\label{plot_pn}

144: \end{center}

145: \end{figure}

146: 

147: Taking into account (\ref{eq:nphotprob}) the expressions for the mean and mean 

148: square values of the registered photon numbers (\ref{eq:meann2}) turn to:

149: 

150: \begin{equation}

151: \label{eq:meann3}

152: \langle N \rangle = \frac{|\lambda|^2 I_1(2|\lambda|)}{I_0 (2|\lambda|)} , 

153: \left\langle {N^2 } \right\rangle = \left| \lambda \right|^2

154: \end{equation}

155: 

156: 

157: The measurements have the statistical uncertainty, caused by quantum 

158: fluctuations. This uncertainty can be characterized 

159: by the corresponding dispersion:

160: 

161: \begin{equation}

162: \sigma ^2 = \langle {N^2} \rangle - \langle 

163: {N} \rangle ^2

164: \end{equation}

165: 

166: 

167: Taking into account (\ref{eq:meann3}) we get the following 

168: expression:

169: 

170: \begin{equation}

171: \label{eq:sigma}

172: \sigma ^2 = \left| \lambda \right|^2\left( {1 - \left( {\frac{I_1 

173: \left( {2\left| \lambda \right|} \right)}{I_0 \left( {2\left| \lambda 

174: \right|} \right)}} \right)^2} \right)

175: \end{equation}

176: 

177: The dependencies of the measurement results uncertainty on the mean photon number for

178: the TMCC-beam and a usual correlated beam are given at \fref{plot_dispers}. 

179: The difference between these dependencies can also be used for the TMCC-states identification.

180: 

181: \begin{figure}[htbp]

182: \begin{center}

183: 	\epsfbox{plot_dispers.eps}

184: 	\caption{The dependencies of the measurement results uncertainty on the mean photon number for the TMCC-beam (solid line) and a usual coherent beam (dotted line)}

185: 	\label{plot_dispers}

186: \end{center}

187: \end{figure}

188: 

189: \section{Communication via quantum channel}

190: 

191: Let we have to establish a secure quantum channel between two parties (\fref{scheme1}). Alice has the laser on her side, which produces two beams in the TMCC 

192: state. The optical channel is organized in such a way, that Alice receives one 

193: of the modes, the first, for example, i.e. $\varphi _A \equiv \varphi _1 

194: $,$\varphi _A (x_A ,t_0 ) = 1$ , and Bob receives another one, i.e. $\varphi _B 

195: \equiv \varphi _2 $ ,$\varphi _B (x_B ,t_0 ) = 1$ at any moment of 

196: measurement $t_0 $, where $x_A $and $x_B $are Alice's and Bob's locations 

197: respectively. Accordingly, Alice cannot measure the Bob's beam and vice 

198: versa:$\varphi _B (x_A ,t_0 ) = 0$, $\varphi _A (x_B ,t_0 ) = 0$. At that 

199: the vector-potential of the TMCC-beam is:

200: 

201: \begin{equation}

202: A = \varphi _A^\ast (x,t)a_A^ + + \varphi _A (x,t)a_A + \varphi _B^\ast 

203: (x,t)a_B^ + + \varphi _B (x,t)a_B 

204: \end{equation}

205: 

206: 

207: Unlike the usual non-correlated coherent states, which

208: show their quasiclassical properties in the fact, that the mean value of a vector-potential of

209: a corresponding beam is not equal to 0, the mean value of a vector-potential of a TMCC-beam 

210: and any other characteristic, which is linear in field, turns to be equal to 0, because during 

211: the averaging by the 

212: 1$^{st}$ mode, for example, the $a_1$ converts $\left| {n,n} \right\rangle $ to $\left| 

213: {n - 1,n} \right\rangle $, which is orthogonal to all the present state 

214: terms, so $\left\langle {\lambda _i } \right|a_i \left| {\lambda _i } 

215: \right\rangle = 0$. So the quasiclassical properties in their usual meaning are absent in 

216: the case of a TMCC-beam. But they become apparent in the non-zero value of the spatial correlation function,

217: which characterizes the interdependence of the results of measurements 

218: taken by Alice and Bob:

219: 

220: 

221: \begin{equation}

222: g_{AB} = < N_A N_B > - < N_A > < N_B > 

223: \end{equation}

224: 

225: \begin{figure}[htbp]

226: 	\epsfbox{scheme1.eps}

227: 	\caption{Quantum channel between two parties with a TMCC source}

228: 	\label{scheme1}

229: \end{figure}

230: 

231: 

232: 

233: It's useful to describe the channel quality by the relative correlation, 

234: which is

235: 

236: \begin{equation}

237: \rho _{AB} = \frac{ < N_A N_B > - < N_A > < N_B > }{\sigma _A \sigma _B }

238: \end{equation}

239: 

240: 

241: The main feature of the TMCC state is that the value $\rho _{AB} $ is 

242: exactly equal to 1, while in the case of non-correlated beams we would get 

243: $\rho _{AB} = 0$. This means that the measurements of the photon numbers, 

244: obtained by Alice and Bob, each with her/his own detector, not only show the same 

245: mean values, but

246:  even have the same deflection from the mean values. 

247: 

248: 

249: 

250: The laser beam is the semi-classical radiation with well defined phase, but 

251: due to the uncertainty principle for the number of photons and the 

252: phase of the radiation, there is a large enough uncertainty in the photon 

253: numbers, this can be seen from the dispersion expression (\ref{eq:sigma}). Thus one can 

254: observe the noise, which is similar to the shot noise in an electron tube. 

255: In the TMCC radiation the characteristics of such noise for each of the 

256: modes are amazingly well correlated to each other. This fact 

257: enables the use of such radiation for generation of a random code, which 

258: will be equally good received by two mutually remote detectors. 

259: 

260: 

261: \subsection{The protocol}

262: 

263: 

264: The following scheme can be used for the TMCC-based protocol. The laser is 

265: set up to produce the constant mean number of photons during the session and 

266: both parties know this number. At some moment Alice and Bob start the 

267: measurements. They detect the number of photons at unit time  by measuring 

268: the integrated intensity of the corresponding incoming beam. If the number 

269: of photons for the specific unit time is larger than the known expected mean 

270: (which is due to the shot noise), the next bit of the generated code is 

271: considered to have the value ``1''. If the measured number is less than 

272: the expected mean, the next bit is considered to be equal to ``0'':

273:  

274: \begin{equation}

275: \label{protocol}

276: B ={\left\{

277: \begin{array}{l}

278: \{n \leq [<N>]\} \rightarrow 0 \\ \{n > [<N>]\} \rightarrow 1

279: \end{array}\right.}

280: \end{equation}

281:  

282:  

283: Upon the receipt of a sufficient number of bits (the code), both Bob and Alice 

284: divide them in half, each obtaining two bit sequences (half-codes). Bob encodes one 

285: sequence with another, using the "eXclusive OR" logical operation (XOR, $B_i\bigoplus B_j$), 

286: and sends this encoded half-code to Alice using any public channel. 

287: Alice uses any of her half-codes to

288: decode the code she has got from Bob using the same XOR operation. She compares

289: the result of this operation with another of her half-codes. If

290: all the bits coincide, this means that Alice and Bob both have the same

291: code, which can be used as a cryptographic key for encoding their communication.

292: Otherwise they have to repeat the key generation and transfer procedure and check 

293: the channel for the possible eavesdropping if the procedure fails again.

294: 

295: The stability of the protocol against the basic beam splitting attacks was

296: examined in \cite{tmcc} and it was shown that any successful

297: attempt destroys the channel and cancels the key distribution session.

298: Besides the basic listening-in, Eve may carry out a more advanced

299: state cloning eavesdropping attack by detecting the overall photon

300: number in the Bob's mode for each upcoming bit and then re-emiting the

301: same number again, which requires Eve to have the same laser source on her site.

302: Though, due to the laser shot noise Eve can't be sure if she is

303: producing exactly the same number of photons, she can set up her

304: laser just to produce some mean photon number. In case she will

305: adjust her laser to produce the mean photon number, which is equal to

306: the current photon number measured in the Bob's mode for a next

307: bit, she may probably be successfull by repeating some bits. But in

308: this case she will change the Bob's measurement results ditribution.

309: This can be checked by expanding a protocol with a post-measurement

310: analysis of the measurement results, which can be done by both of

311: the trusted parties. It should consist of the comparison of an

312: obtained photon numbers distribution by the frequencies of their

313: detection with the expected one for the known constant mean photon number

314: value which is actually a task of comparing two numerical arrays

315: The difference between expected and obtained distributions will

316: reveal a state cloning attempt.

317: 

318: \subsection{Quantum channel error analysis}

319: 

320: 

321: Let the parties of the secret key transmission procedure are using the protocol 

322: described above, thus they estimate the value of the next bit by comparing the 

323: actual registered photon number to the average. The probability of detecting "0" 

324: bit value then is

325: 

326: \begin{equation}

327: P_{(0)} (\lambda) = \sum_{n=0}^{[\langle N \rangle]}P_n(\lambda)

328: \end{equation}

329: 

330: The noise is present in the channel and it may increase the number of the registered 

331: photons. We suppose that the noise is thermal and assume that 

332: it may, with some probability, cause an appearance of one and no more than one 

333: additional photon in any of the modes during the time of a bit detection. 

334: We will denote the probability of a noise photon detection as $\epsilon$ 

335: and refer to it as the noise factor.

336: 

337: We suppose that the channel is qualitative enough to transfer the impulse at the 

338: required distance without losing any single photon, thus errors are possible only 

339: due to the appearance of the noise photons.

340: 

341: An error, when Alice registers "0" bit value and Bob registers "1" may occur upon 

342: the joint realization of two events. The first is that Alice detects the maximum possible

343: number of these, corresponding to the "0" bit value, which is, according to the proposed protocol, 

344: equal to the integer part of $\langle n \rangle$. The second is that in addition to this number Bob 

345: detects the appearance of a noise photon. The opposite situation, when Alice gets the 

346: "1" bit value and Bob registers "0" is possible when the noise photon was detected 

347: by Alice, the probability of such error is the same. 

348: 

349: The probability of the realization of a state, which consists of the maximum possible 

350: for the "0" bit value photons and, at the same time, is detected as "0", is the 

351: relation between the corresponding probabilities:

352: 

353: \begin{equation}

354: P_{max(0)} = \frac{P_{[\langle N \rangle]}}{P_{(0)}} = \frac{P_{[\langle N \rangle]}}{\sum_{n=0}^{[\langle N \rangle]}P_n(\lambda)}

355: \end{equation}

356: 

357: We will refer to this probability as to the error factor.

358: So the probability of an error during the bit registration is equal to the product 

359: of the noise and error factors:

360: 

361: \begin{equation}

362: P_{err(0)}(\lambda)=\epsilon P_{max(0)} 

363: \end{equation}

364: 

365: One can easily see that upon the intensity increase the error factor becomes less and 

366: so the channel tends to a self-correction if the beam gets more intensive. 

367: 

368: 

369: 

370: 

371: \section{Conclusions}

372: 

373: Correlated coherent states of the two-mode laser beam (TMCC states) show 

374: interesting properties, which can be used, in particular, for the tasks of 

375: the quantum communication and cryptography. 

376: 

377: The TMCC-beams can be identified due to the special form of the registration 

378: probabilities distribution for different photon numbers in the corresponding beam 

379: and the dependence of the dispersion on the mean photon numbers value.

380: 

381: On the one hand, each of the 

382: modes looks like a flow of the independent photons rather then a coherent 

383: beam, since mean values of the operators, which are linear in field, are 

384: equal to 0 for each mode separately. 

385: 

386: On the other hand, the strong 

387: correlation between the results of measurements for each of the modes takes 

388: place. This correlation shows itself in the fact that in each of the modes numbers 

389: of photons are the same and even the shot noise shows itself equally in the both 

390: modes. This enables the use of the TMCC state as the generator and carrier 

391: of random keys in a quantum channel which is stable against the eavesdropping \cite{tmcc}. 

392: 

393: 

394: Thus, the TMCC-laser generates and transmits exactly the 2 copies of a 

395: random key. Unlike the single or two-photon schemes, which require large numbers of 

396: transmission reiterations to obtain the statistically significant 

397: results, the TMCC beam can be intensive enough to make each single measurement statistically 

398: significant and thus to use single impulse for each 

399: piece of information, and remain cryptographically steady. This allows to essentially increase 

400: the effective data transfer rate and distance. Analysis of the noise influence on

401: the channel properties shows that the channel tends to a self-correction upon the beam intensity increase.

402: 

403: 

404: \Bibliography{99}

405: \bibitem{qc1} Nicolas Gisin, Gregoire Ribordy, Wolfgang Tittel, Hugo Zbinden. Quantum Cryptography. Preprint: quant-ph/0101098

406: 

407: \bibitem{qc2} Matthias Christandl, Renato Renner, Artur Ekert.  A Generic Security Proof for Quantum Key Distribution. Preprint: quant-ph/0402131

408: 

409: \bibitem{qc3} Nicolas Gisin, Nicolas Brunner. Quantum cryptography with and without entanglement. Preprint: quant-ph/0312011

410: 

411: \bibitem{per} A. Perelomov, Generalized Coherent States and Their Applications (Springer, Berlin, 1986).  

412:  

413: \bibitem{ent1} Wolfgang Tittel, Gregor Weihs. Photonic Entanglement for Fundamental Tests and Quantum Communication. quant-ph/0107156

414: 

415: \bibitem{tmcc} Constantin V. Usenko and Vladyslav C. Usenko. Preprint: quant-ph/0403112 (submitted to Journal of Russian Laser Research)

416: 

417: \bibitem{entprot} A. Ekert, Phys. Rev. Lett. 67, 661 (1991) 

418: 

419: \bibitem{entprotexp} D. S. Naik et al., Phys. Rev. Lett. 84, 4732 (2000)

420: 

421: \bibitem{single1} C. H. Bennett, Phys. Rev. Lett. 68, 3121 (1992)

422: 

423: \bibitem{single2} C. K. Hong and L. Mandel, Phys. Rev. Lett. 56, 58 (1986)

424: 

425: \bibitem{four} C. H. Bennett and G. Brassard , “Quantum cryptography: public key distribution and coin tossing”, Int . conf. Computers, Systems \& Signal Processing, Bangalore, India, 1984, 175- 179.

426: 

427: \bibitem{fourexp} T. Jennewein et al., Phys. Rev. Lett. 84, 4729 (2000)

428: 

429: \bibitem{polar} A.C. Funk, M.G. Raymer. Quantum key distribution using non-classical photon number cor\-re\-la\-tions in macroscopic light pulses. Preprint: quant-ph/0109071

430: \bibitem{similar1} Yun Zhang, Katsuyuki Kasai, Kazuhiro Hayasaka. Quantum channel using photon number correlated twin beams. quant-ph/0401033, Optics, Express 11, 3592 (2003)

431: \bibitem{similar2} L. A. Wu, H. J. Kimble, J. L. Hall, and H. F. Wu, “Generation of squeezed states by parametric down conversion,” Phys. Rev. Lett. 57, 2520-2524 (1986). 

432: \bibitem{similar3} H. Wang, Y. Zhang, Q. Pan, H. Su, A. Porzio, C. D. Xie, and K. C. Peng, “Experimental realization of a quantum measurement for intensity difference fluctuation using a beam splitter,” Phys. Rev. Lett. 82, 1414-1417 (1999). 

433: 

434: \endbib

435: 

436: \end{document}

437: