1: \documentclass[pra,aps,showpacs,epsfig,amsmath,amsthm,amsfonts,amssymb,twocolumn]{revtex4}
2:
3: \usepackage{graphicx}% Include figure files
4:
5: \begin{document}
6: \title{Quantum Interference in the Kirkwood-Rihaczek representation}
7: \author{L. Praxmeyer and K. W\'odkiewicz}
8: \affiliation{Institute of Theoretical Physics,
9: Warsaw University,\\
10: Ho\.za 69, 00--681 Warsaw, Poland }
11:
12: \date{10 July 2002 }
13:
14: \begin{abstract}
15: We discuss the Kirkwood-Rihaczek phase space distribution
16: and analyze a whole new class of quasi-distributions connected with this
17: function. All these functions have the correct marginals.
18: We construct a coherent state representation of such functions, discuss
19: which operator ordering corresponds to the Kirkwood-Rihaczek distribution
20: and their generalizations, and show how such states are connected to squeezed
21: states. Quantum interference in the Kirkwood-Rihaczek representation
22: is discussed.
23: \end{abstract}
24:
25: \pacs{PACS number(s): 03.65.Bz, 42.50.Dv}
26:
27: \maketitle
28:
29:
30: \section{Introduction}
31: A number of different phase space distribution functions has been
32: introduced and investigated over the years, the Wigner
33: distribution function \cite{Wigner} being the most famous and the
34: most known of all. The phase space distributions from the
35: Glauber-Cahill s-parameterized class of quasi-distributions
36: \cite{GlaCah} that contains the Wigner function, the
37: Glauber-Sudarshan $P$-representation \cite{Gla,Sud}, and the
38: Husimi or the $Q$-representation \cite{Husimi}, have been widely used
39: as useful and powerful phase space tools.
40:
41:
42: It is the purpose of this paper to show that a different class of phase space
43: distribution functions,
44: connected with a lesser-known Kirkwood distribution function
45: \cite{Kirkwood}, can be useful in the investigation of quantum
46: interference in phase space. The Kirkwood
47: distribution was proposed just a year after Wigner introduced his
48: function and like the Wigner distribution was firstly used in
49: quantum statistics and thermodynamics. The Kirkwood distribution
50: has been rediscovered by Rihaczek \cite{Rihaczek} for use in the
51: theory of time-frequency analysis of classical signals. Operator
52: bases of Kirkwood and Wigner distribution functions have been
53: examined by Englert \cite{Englert}. Zak \cite{Zak} studied the
54: Rihaczek function in the quasimomentum-quasicoordinate
55: representation.
56:
57: In this paper we shall study a new class of phase space distribution functions
58: connected with the Kirkwood-Rihaczek distribution function.
59: The main advantage of the functions from this class is that they all lead to
60: the correct position and momentum marginals. We shall show how to
61: construct a coherent state representation of such functions and which
62: operator ordering corresponds to the Kirkwood-Rihaczek
63: distribution and their generalizations.
64: A connection of such phase space
65: functions with squeezed states will be established.
66:
67:
68: In Section \ref{sec2} we present the basic definitions, properties, and
69: differences
70: of the Wigner and Kirkwood distributions. Section \ref{sec3} is devoted to the problem of operator ordering and its relation to various phase space
71: distributions. In Section \ref{sec4} the definition of
72: generalized Kirkwood-Rihaczek distribution function is introduced and it is
73: shown that Kirkwood-Rihaczek function fully characterize the quantum state.
74: This means that
75: a full quantum state reconstruction with the help of the Kirkwood-Rihaczek function is possible.
76: Sections \ref{sec5} and \ref{sec6} are devoted
77: to the quantum interference in phase space. Using simple examples, we present
78: various properties of the Kirkwood-Rihaczek functions and we discuss similarities and differences of such phase space quasi-distributions with the Wigner function.
79:
80: \section{Phase space and quantum marginals}\label{sec2}
81: \subsection{The Wigner distribution function}
82: In 1932 Wigner introduced a phase-space distribution function
83: \begin{equation}
84: W_{\Psi}(q,p)=
85: \frac{1}{2\pi\hbar}\;\int
86: \Psi^{\star}(q+\xi/2)\; e^{\frac{ip\xi}{\hbar}}\;\Psi(q-\xi/2) d\xi\;,
87: \label{zwig}
88: \end{equation}
89: which fulfills the fundamental requirement for a joint probability
90: distribution in phase space, i.e. when integrated over $q$ or $p$, it
91: gives marginal probabilities: $ |\Psi(q)|^{2} $ and $\frac{1}{2\pi\hbar}\;
92: |\widetilde{\Psi}(p)|^{2} $,
93: where
94: $
95: \widetilde{\Psi}(p)=\int dq\, e^{-\frac{ipq}{\hbar}}\,\Psi(q)\;
96: $
97: is the Fourier transform of $\Psi(q)$.
98:
99: As shown by Wigner, the function given by Eq. (\ref{zwig}) is in
100: general not positive. However, under simple and reasonable
101: physical
102: assumptions it is unique. Thus, when one demands that:
103: (i) a phase space distribution $P(q,p)$ is real, (ii) bilinear in $\Psi$, (iii)
104: gives the correct marginals, and (iv) its dynamical evolution reproduces
105: the Liouville equation in the classical limit, then the distribution $P(q,p)$
106: is unique and is the Wigner function \cite{O'Connell}.
107:
108: After Wigner's work many different phase space
109: distribution functions have been introduced and investigated. There is a very rich
110: literature devoted to applications of the Wigner function and other various
111: quasi-distributions in quantum optics (see e.g. \cite{ff}).
112:
113: The simplest example of a phase space distribution which does not satisfy the
114: four assumptions leading to the Wigner function, but reproduces the right
115: marginal in an explicit way, is the distribution function:
116: \begin{equation}
117: P(q,p)=\frac{1}{2\pi\hbar} |\Psi(q)|^{2}\;
118: |\widetilde{\Psi}(p) |^{2} .
119: \label{prop}
120: \end{equation}
121: Clearly, this distribution is not bilinear in $\Psi$ and as a result does not
122: satisfy the requirements of the Wigner's uniqueness theorem. Another
123: distribution function similar to that given by Eq. (\ref{prop}) can be
124: guessed taking formally a
125: square-root of this expression. As a result, up to an arbitrary phase, we have
126: \begin{equation}
127: P(q,p)\sim |\Psi(q)|\;
128: |\widetilde{\Psi}(p) | .
129: \label{prop1}
130: \end{equation}
131: Note that this distribution function contains no information about
132: the phase of the wave function. A simple insertion of an
133: additional phase factor $\varphi(q,p)$ to the wave functions leads
134: to the expression:
135: \begin{equation}
136: P(q,p)\sim \Psi(q) \; e^{i\varphi(q,p)}\; \widetilde{\Psi}^{\star}(p)\,,
137: \label{prop2}
138: \end{equation}
139: that defines a class of bilinear but complex distribution
140: functions. An example of such a distribution has been proposed
141: just one year after Wigner introduced his famous distribution
142: function.
143:
144: \subsection{The Kirkwood-Rihaczek distribution function}
145: In 1933 Kirkwood introduced
146: a phase space distribution which, according to his description, ``{\em \ldots
147: differs but little from the Wigner function}''\cite{Kirkwood}. The Kirkwood function is defined
148: as follows:
149: \begin{eqnarray}
150: K(q,p)&=& \frac{1}{2\pi\hbar} \int d\xi\, \Psi(q)\,
151: e^{\frac{i(\xi-q)p}{\hbar}}\,\Psi^{\star}(\xi)=\nonumber\\
152: & =&\frac{1}{2\pi\hbar}\, \Psi(q)\, e^{- \frac{ipq}{\hbar}}\,
153: \widetilde{\Psi}^{\star}(p).
154: \label{zrih}
155: \end{eqnarray}
156: In 1968, this function was rediscovered by Rihaczek in the context of a signal
157: energy distribution in time and frequency (for a review of time-frequency
158: distributions see \cite{Cohen0,Coh01}).
159:
160: The real part of the Kirkwood-Rihaczek distribution
161: \begin{equation}
162: K_{{\mathrm{Re}}}(q,p)={\mathrm{Re}}\left[ \Psi(q)\, e^{- \frac{ipq}{\hbar}}\,
163: \widetilde{\Psi}^{\star}(p)\right] \label{re-zrih}
164: \end{equation}
165: is closely related to a quantum mechanical phase space distribution
166: introduced by Margenau and Hill \cite{M-H}.
167:
168: We clearly see that the Kirkwood-Rihaczek (K--R) distribution has the correct
169: marginal properties:
170: \begin{equation}
171: \int K(q,p)\,dp\, =\, |\Psi(q)|^{2}\;,\nonumber
172: \end{equation}
173: \begin{equation}
174: \int K(q,p)\,dq\, =\,\frac{1}{2\pi\hbar}\;
175: |\widetilde{\Psi}(p)|^{2}\;,\label{kweew}
176: \end{equation}
177: and it is normalized,
178: \begin{equation}
179: \iint K(q,p)\, dq\,dp=1,
180: \end{equation}
181: as a consequence of the normalization of the wave function
182: $\Psi(q)$. Note that the contribution of the imaginary part of the
183: K--R distribution to the marginals is identically zero. Hence, we
184: shall often focus on the real part.
185:
186: Another condition easy to find is that the absolute square of
187: $K(q,p)$ has the form similar to the Eq.(\ref{prop}), i.e.
188: \begin{equation}
189: \mid K(q,p)\mid^2\,= \frac{1}{(2\pi\hbar)^2}\, |\Psi(q)|^{2}\;
190: |\widetilde{\Psi}(p) |^{2},\label{absq}
191: \end{equation}
192: which indicates that $
193: K(q,p)$ is a
194: square-integrable function, with
195: \begin{equation}
196: \iint \mid K(q,p)\mid^2 dq\,dp=\frac{1}{(2\pi\hbar)}.
197: \end{equation}
198: The absolute square of the K--R function, Eq. (\ref{absq}), has a simple
199: physical interpretation. It is just proportional to the product of the
200: probabilities in configuration and momentum representations.
201:
202: The dynamical free evolution of a particle with mass $m$ of the K--R
203: distribution function is given by the following equation
204: \begin{equation}
205: \partial_{t} K(q,p,t)\;+\;
206: \frac{p}{m}\;\partial_{q}
207: K(q,p,t)=\frac{i\hbar}{2m} \partial_{q}^2
208: K(q,p,t)\; ,
209: \label{ewol:rih} \end{equation}
210: which can be also written in a form:
211: \begin{equation*}
212: K(q,p,t)=e^{t(
213: \frac{i\hbar}{2m}\partial_{q}^2-
214: \frac{p}{m}\partial_{q})}K(q,p,0).
215: \end{equation*}
216: We see from this formula, that the free evolution of the K--R
217: function is a superposition of the free Schr\"odinger diffusion
218: and of a classical boost to a moving frame. The K--R distribution
219: function is bilinear and has the correct marginal properties but,
220: in contrast with the Wigner function, it is not real nor does its
221: free evolution satisfy the classical Liouville equation. As we
222: shall see in detail later, it is also not well-behaved under
223: rotations of the $(p,q)$ coordinate system, and this implies that
224: such a function can not be measured by tomography methods.
225:
226: The simplicity of the definition, Eq. (\ref{zrih}), indicates that it is
227: relatively easy to evaluate the K--R distribution function even
228: for systems for which an analytical formula of the Wigner function
229: is not known. The best example of such a system is a hydrogen
230: atom. Elsewhere we shall present the K--R functions for different
231: energy levels of the hydrogen atom \cite{new}.
232: \subsection{The Cohen distribution functions}
233: We have already emphasized the marginal properties of a phase space
234: distribution
235: function.
236: % relaxing for example the reality of such a function,
237: An interesting question arises about a general form of the distribution with
238: the correct marginals. This problem has been posed and solved by Cohen in 1966
239: \cite{Cohen,Cohen2}. The most general distribution $P(q,p)$ with the proper
240: marginals has the form of a double Fourier transform of a function
241: \begin{equation}
242: A(q',p') = \int e^{-ip'\xi/\hbar}\;
243: \Psi^{\star}(\xi-q'/{2})\;\Psi(\xi+q'/{2})\;d\xi\; \label{zamb}
244: \end{equation}
245: multiplied by an arbitrary function $\Phi(q',p')$ satisfying the
246: relations
247: \begin{equation}
248: \Phi(q',0)=1=\Phi(0,p')\;.
249: \label{war}
250: \end{equation}
251: In the literature devoted to optical processing of classical signals, the
252: function $A(q',p')$ is called the Ambiguity function.
253:
254:
255: Therefore, the Cohen joint distribution functions labeled by
256: functions
257: $\Phi$ are given by the following equation
258: \begin{eqnarray}
259: & & P_{\Phi}(q,p)={\cal F}\, [\,\Phi\, A\,]:=\label{uog:wig}\\
260: & &=\frac{1}{(2\pi\hbar)^{2}} \iint e^{i(p'q-q'p)/\hbar}
261: \Phi(q',p')A(q',p')dp'dq',\nonumber
262: \end{eqnarray}
263: where by ${\cal F}$ we have denoted a double Fourier transform.
264:
265: The Wigner distribution function is obtained by substituting $\Phi(q',p')=1$
266: in the Cohen
267: distribution functions formula. The K--R distribution function is obtained for
268: $\Phi(q',p')=\exp[-ip'q'/(2\hbar)]$;
269: $\Phi(q',p')=\cos[p'q'/(2\hbar)]$ leads to the Margenau-Hill distribution.
270:
271: \section{Wigner-Weyl transformation}\label{sec3}
272: \subsection{The $(q,p)$ phase space description}
273: The Wigner-Weyl association of classical phase space
274: functions ${\cal A}(q,p)$ with quantum
275: operators follows form the property:
276: \begin{eqnarray}
277: & &\{ {\cal A}(\hat{q},\hat{p})\}_{\rm ordering}=\\
278: & & =\iint dq dp\,
279: {\cal A}(q,p)\,\{ \delta(q-\hat{q})\delta(p-\hat{p}) \}_{\rm ordering}.
280: \nonumber
281: \end{eqnarray}
282: The problem of defining a quantum operator in phase space lies in the ordering of operators $\hat{q} $ and $
283: \hat{p}$. As an example of such association we shall take a phase space
284: density distribution, whose quantum average leads to the density operator of
285: the system.
286: Using the Fourier decomposition of the Dirac delta functions we see
287: that the ordering becomes equivalent to an ordering of the Heisenberg-Weyl
288: algebra operators:
289: \begin{eqnarray}
290: & &\{\delta(q-\hat{q})\delta(p-\hat{p}) \}_{\rm ordering} =\label{dirac}\\
291: & &=\frac{1}{(2\pi\hbar)^2}
292: \iint dq' dp'
293: e^{\frac{i(p'q-q'p)}{\hbar}}
294: \left\{ e^{\frac{i\hat{p}q'}{\hbar} }
295: e^{-\frac{ip'\hat{q}}{\hbar} } \right\}_{\rm ordering}.\nonumber
296: \end{eqnarray}
297: This formula shows that there is no natural unique generalization of the
298: classical probability density in phase space, because it is possible to have
299: arbitrary classes of operator orderings. It has been recognized by Wigner and
300: Weyl that various quantum distribution functions can be associated with
301: different orderings of operators $\hat{q}$ and $\hat{p}$.
302: For example,
303: the Wigner distribution function corresponds to Wigner-Weyl ordering, which is
304: obtained by putting $\hat{q} $ and $ \hat{p}$ operators in the same exponent in
305: Eq. (\ref{dirac}). We shall show that the K--R distribution function
306: corresponds to a special ordering called the anti-standard ordering.
307:
308: One can easily find that the Ambiguity function, Eq.(\ref{zamb}), may be expressed also as
309: \begin{equation}
310: A(q',p')= \langle \Psi| e^{\frac{i\hat{p}q'-ip'\hat{q}}{\hbar}}
311: |\Psi\rangle\;.
312: \label{a1}
313: \end{equation}
314: Using the above equation, the definition of the K--R distribution function can
315: be formulated as follows
316: \begin{equation}
317: K(q,p)
318: =\frac{1}{(2\pi\hbar)^{2}}
319: \iint dp'dq' e^{\frac{i(p'q-q'p)}{\hbar}}
320: \langle \Psi| e^{\frac{i\hat{p}q'}{\hbar}}
321: e^{\frac{-ip'\hat{q}}{\hbar}}
322: |\Psi\rangle\;.
323: \nonumber
324: \end{equation}
325: By comparison with Eq. (\ref{dirac}) we find that the Kirkwood distribution
326: function corresponds to an ordering such that all $\hat{p} $ operators are on the left of all $\hat{q}$ operators. Such
327: an arrangement of the canonical operators is called the anti-standard ordering.
328: Analogously one can find that the complex conjugation of the K--R function
329: corresponds to the standard ordering (all operators $ \hat{q}$ are on the left
330: followed by all $ \hat{p}$ operators). The real part of K--R distribution
331: function is associated with a symmetric superposition of the anti-standard and
332: the standard ordering, i.e.
333: $\frac{1}{2}(\hat{q}^{n}\hat{p}^{m}+\hat{p}^{m}\hat{q}^{n})$. Note that such
334: an ordering is not equivalent to the Wigner-Weyl ordering leading to the Wigner
335: function.
336:
337: \subsection{Coherent state phase space description}
338: So far various properties and definitions of phase space
339: distribution
340: functions were formulated in position or momentum representations.
341: Such a parameterization seems to be the most natural to study phase space
342: properties of particles. However, Glauber and Cahill have pointed out that the
343: coherent state representation is more natural and useful while dealing with
344: phase space distribution functions describing the quantum states of light
345: \cite{GlaCah}.
346:
347: We shall now use the following notation:
348: Greek letters ($\alpha,\; \beta$, etc.) designate complex variables;
349: $\hat{a}$, $\hat{a}^{\dagger}$ annihilation
350: and creation operators;
351: $\mathrm{\hat{D}}(\alpha)=
352: \exp(\alpha \hat{a}^{\dagger} -\alpha^{\star} \hat{a})$
353: denotes displacement operator; and all integrations are taken over
354: the whole complex plane. Using the formalism of coherent states
355: Glauber and Cahill have shown that it is possible to define an
356: $s$-parameterized class of distribution functions simply related
357: to the Wigner distribution function. These quasi-distributions are
358: defined as a complex Fourier transform ${\cal F}$
359: of the $s$-ordered characteristic function:
360: \begin{equation}
361: W(\alpha,s)={\cal F}\,[\,C\,] :=\int \frac{d^{2}\beta}{\pi^{2}}\;
362: e^{\alpha\beta^\star-\alpha^\star\beta} C(\beta,s),
363: \label{wigs}
364: \end{equation}
365: defined as
366: \begin{equation}
367: C(\beta,s)=e^{\frac{s-1}{2} |\beta|^{2}}
368: \mathrm{Tr}\left[\hat{\rho}
369: e^{\beta a^{\dagger}}e^{-\beta^{\star} a}\right]=
370: e^{\frac{s}{2} |\beta|^{2}}
371: \mathrm{Tr}[\hat{\rho}\mathrm{\hat{D}}(\beta)].
372: \end{equation}
373: In the equation above $\hat{\rho}$ is the density operator of the investigated system. The normally ordered form of
374: $\mathrm{\hat{D}}(\beta)$ allows to perform the integral explicitly, and then Eq.
375: (\ref{wigs}) can be rewritten as
376: \begin{equation}
377: W(\alpha,s) = \frac{2}{\pi(1-s)} \mathrm{Tr}
378: \left[ \hat{\rho} \mathrm{\hat{D}}(\alpha)
379: \hat{\Pi}(s)
380: \mathrm{\hat{D}}^{\dagger}(\alpha)\right]
381: \label{wig:hatn}
382: \end{equation}
383: with the operator $\hat{\Pi}(s)$ defined as:
384: \begin{equation}
385: \hat{\Pi}(s)=\left( \frac{s+1}{s-1} \right)^{\hat{a}^{\dagger}\hat{a}}.
386: \end{equation}
387: The continuous parameter $s$ (which is required to be real and to satisfy an
388: inequality $s \leq 1$) corresponds to differing ordering of the creation and
389: annihilation operators. Three values: $s=1$, $s=0$, and $s=-1$ correspond to
390: normal, symmetric, and anti-normal ordering, that lead to the Glauber-Sudarshan
391: $P$-representation,
392: % \cite{Sud,Glau},
393: the Wigner function, and the $Q$-representation, respectively.
394:
395: These results have been generalized by Agarwal and Wolf in \cite{AgW}.
396: They have
397: proposed the following general formula for the quasi-distribution functions:
398: \begin{equation}
399: W(\alpha,\Omega)=\int\frac{d^{2}\beta}{\pi^{2}}
400: e^{\alpha\beta^\star-\alpha^\star\beta}
401: \Omega(\beta,\beta^{\star})
402: \mathrm{Tr}\left[\hat{\rho} \mathrm{\hat{D}}(\beta)\right] \;,
403: \label{Omeg}
404: \end{equation}
405: where
406: $\Omega(\beta,\beta^{\star})$ is an analytic function of the complex variable
407: $\beta$ that has no zeros.
408: The condition $\Omega(0,0)=1 $ guarantees the normalization
409: \begin{equation}
410: \int d^2\alpha\, W(\alpha,\Omega)=1.
411: \end{equation}
412: Obviously, such functions do not provide automatically the correct quantum
413: marginals, and only for very selected functions from the class given by Eq.
414: (\ref{Omeg}), we can satisfy the relations from Eq. (\ref{kweew}).
415:
416: In the following Section we shall show that the K--R distribution
417: corresponds to a specific form of the $\Omega$ function. Moreover,
418: with the help of this formalism we shall define a new class of
419: $\sigma$-ordered K--R distribution functions in full analogy to
420: the $s$-ordered functions of Cahill and Glauber.
421:
422: \section{Generalized K--R function and quadratic ordering}\label{sec4}
423: \subsection{Definition of generalized K--R distribution}
424: A new class of quasi-distributions involving a quadratic ordering function:
425: $\Omega(\beta,\beta^{\star})=\exp\left(\sigma(\beta^{\star\;2}-
426: \beta^{2})/4 \right) $
427: is defined as a Fourier transform
428: \begin{equation}
429: K(\alpha,\sigma)= {\cal F}[C(\beta,\sigma)]=\int \frac{d^{2}\beta}{\pi^{2}}
430: e^{\alpha\beta^\star-\alpha^\star\beta} C(\beta,\sigma), \label{k2}
431: \end{equation}
432: of the characteristic function given by
433: \begin{equation}
434: C(\beta,\sigma)=e^{\sigma\frac{\beta^{\star\;2}-\beta^{2}}{4}}
435: \mathrm{Tr}(\hat{\rho} \mathrm{\hat{D}}(\beta)) \,.
436: \label{RIH1}
437: \end{equation}
438: The continuous parameter $\sigma$ is real and there are no limits to its value.
439: The K--R distribution function corresponds to $\sigma=1$, for $\sigma=0$ we
440: again obtain the Wigner distribution function.
441:
442:
443: As in the case of s-ordered Cahill-Glauber
444: distribution functions, we can perform the integrals and then rewrite
445: Eq. (\ref{k2}) as
446: \begin{equation}
447: K(\alpha,\sigma) =\frac{2}{\pi\sqrt{1+{\sigma}^2}}
448: \mathrm{Tr \left[ \hat{\rho}\;\hat{D}(\alpha)
449: \hat{K}(\sigma)
450: \hat{D}^{\dagger}(\alpha) \right]}, \label{kkir}
451: \end{equation}
452: where
453: \begin{equation}
454: \mathrm{ \hat{K}(\sigma) =\,:\,e^{\frac{-2\hat{a}\hat{a}^{\dagger}+
455: \sigma(\hat{a}^{\dagger}\hat{a}^{\dagger}-\hat{a}\hat{a})}{1+{\sigma}^2}}}\,:\,.
456: \label{oper}
457: \end{equation}
458: In the above formula $\;:\;\;:\;$ denotes normal ordering of operators
459: $\hat{a}$, $\hat{a}^{\dagger}$. Using the identity $t^{\hat{a}^{\dagger}\hat{a}}=
460: \,:e^{(t-1 )\hat{a}^{\dagger}\hat{a}}:\, $ we can transform Eq. (\ref{oper}) into
461: \begin{equation}
462: \mathrm{} \hat{K}(\sigma) =
463: e^{\frac{\sigma\hat{a}^{\dagger}\hat{a}^{\dagger}}{1+{\sigma}^2}}
464: \left(\frac{{\sigma}^2 - 1}{1+{\sigma}^2} \right)^{\hat{a}^{\dagger} \hat{a}}
465: e^{-\frac{\sigma\hat{a}\hat{a}}{1+{\sigma}^2}}\;.\label{opkir}
466: \end{equation}
467:
468: Equations (\ref{kkir}--\ref{oper}) define a $\sigma$-ordered class
469: of generalized K--R phase space distributions. From the property
470: $K^{\dagger}(\alpha,\sigma)=K(\alpha,-\sigma)$, it is clear that
471: the only real function in this class is the Wigner function which
472: corresponds to $\sigma=0$.
473:
474: The operator $\mathrm{\hat{K}}(\sigma) $ is not Hermitian but has a finite trace
475: \begin{equation}
476: \mathrm{Tr} \left[ \mathrm{ \hat{K}}(\sigma)\right]=\frac{\pi}{2}\sqrt{1+{\sigma}^2}.
477: \end{equation}
478: Let us define operators ${\cal\hat{K}}(\alpha,\sigma)$ as
479: \begin{equation}
480: {\cal\hat{K}}(\alpha,\sigma)=\mathrm{\hat{D}(\alpha)
481: \hat{K}(\sigma)
482: \hat{D}}^{\dagger}(\alpha),
483: \end{equation}
484: and study its properties.
485: We find that sets ${\cal\hat{K}}(\alpha,1)$ and
486: ${\cal\hat{K}}^{\dagger}(\alpha,1)$ form a
487: complete, orthogonal basis systems with respect to the scalar product
488: \begin{equation}
489: \mathrm{Tr}\left[{\cal\hat{K}}^{\dagger}(\alpha,1)\,{\cal\hat{K}}(\beta,1)\right]=\frac{\pi}{2}\,
490: \delta(\alpha-\beta).
491: \end{equation}
492: That means that the K--R phase space distribution, up to a normalization
493: factor, is an expansion coefficient of density
494: operator in the basis
495: ${\cal\hat{K}}^{\dagger}(\alpha,1)$.
496: An arbitrary operator $\hat{\rho}$ may be expanded as
497: \begin{equation}
498: \hat{\rho}=\sqrt{2}\int d^{2}\alpha
499: K(\alpha,1){\cal\hat{K}}^{\dagger}(\alpha,1),\label{roz}
500: \end{equation}
501: or, equivalently,
502: \begin{equation}
503: \hat{\rho}
504: =\sqrt{2}\int d^{2}\alpha K(\alpha,-1)\,
505: {\cal \hat{K}}(\alpha,1).
506: \end{equation}
507: From Eq. (\ref{roz}) is clear that knowledge of K--R distribution of the state $\hat{\rho}$
508: is equivalent to the knowledge of the state itself. The same holds for the complex
509: conjugate of K--R distribution function.
510:
511: \subsection{Connection with squeezed states}
512: From the formula given by Eq. (\ref{opkir}) we see that the K--R function
513: corresponds to a special case when the
514: operator $\left(\frac{{\sigma}^2 - 1}{1+{\sigma}^2} \right)^{\hat{a}^{\dagger} \hat{a}} $
515: becomes a projector, as:
516: \begin{equation}
517: \lim _{\sigma\rightarrow 1} \left(\frac{{\sigma}^2 - 1}{1+{\sigma}^2}
518: \right)^{\hat{a}^{\dagger} \hat{a}}=| 0\rangle\langle 0|. \end{equation} Thus,
519: as regards the K--R distribution, the formula given by Eq. (\ref{kkir}) is reduced
520: to
521: \begin{equation}
522: K(\alpha,1)=\frac{\sqrt{2}}{\pi}
523: \mathrm{Tr}\left[\hat{\rho}\;\mathrm{\hat{D}}(\alpha)
524: e^{\frac{\hat{a}^{\dagger}\hat{a}^{\dagger}}{2}}
525: | 0\rangle\langle 0| e^{-\frac{\hat{a}\hat{a}}{2}}
526: \mathrm{\hat{D}}^{\dagger}(\alpha)\right] .
527: \label{rih:0}
528: \end{equation}
529: The above equation may be rewritten in more intuitive form if we refer to
530: the definition of the squeezed state \cite{squeeze}. The general coherent squeezed state
531: $|\alpha,\xi\rangle=\mathrm{\hat{D}}(\alpha)\hat{S}(\xi)|0\rangle$ is obtained by a
532: combined action of the displacement and the squeezing operator
533: \begin{equation}
534: \hat{S}(\xi) =\exp\left(-\frac{\xi}{2}\hat{a}^{\dagger 2}+
535: \frac{\xi^{\star}}{2}\hat{a}^{2}\right)
536: \end{equation}
537: on the vacuum state.
538: The squeezing operator can be written in an ordered form
539: \begin{equation}
540: \hat{S}(\xi) =\exp\left(-\frac{\nu}{2\mu}\hat{a}^{\dagger 2}\right)
541: \left(\frac{1}{\mu}\right)^{\hat{a}^{\dagger}\hat{a}+\frac{1}{2}}
542: \exp\left(\frac{\nu}{2\mu}\hat{a}^{2}\right),\label{squeeze2}
543: \end{equation}
544: where $\mu$ and $\nu$ are functions of squeezing parameter $\xi$
545: given by: $\mu = \cosh{|\xi|}$ and
546: $\nu = e^{i \phi_{\xi} } \sinh{|\xi|}$.
547: Using these relations we find that for $\nu/\mu=1$ and $\phi_{\xi}=0$:
548: \begin{equation}
549: \mathrm{\hat{D}}(\alpha)
550: e^{\frac{\hat{a}^{\dagger}\hat{a}^{\dagger}}{2}}
551: | 0\rangle=\sqrt{\cosh\Gamma}\;|\alpha,-\Gamma\rangle,
552: \end{equation}
553: where $\xi=\Gamma=\arctan$h$(1)=\infty$. This means, that left-hand-side of the
554: equation above corresponds to the infinitely squeezed state. Hence, the
555: definition of the K--R function given by Eq. (\ref{rih:0}) takes the form
556: \begin{equation}
557: K(\alpha,1)=\frac{\sqrt{2}\cosh\Gamma}{\pi}\,
558: \mathrm{Tr}\left[\rho |\alpha,-\Gamma\rangle
559: \langle\Gamma,\alpha|\right].
560: \end{equation}
561: In the similar way we can represent the real part of K--R distribution as
562: \begin{eqnarray}
563: \mathrm{Re}[K(\alpha,1)]=\frac{\sqrt{2}\cosh\Gamma}{2\pi}\,
564: &\mathrm{Tr}& \left[\rho (|\alpha,-\Gamma\rangle
565: \langle\Gamma,\alpha| \right.\nonumber\\
566: &+&\left. |\alpha,\Gamma\rangle
567: \langle -\Gamma,\alpha|) \right].
568: \end{eqnarray}
569: This formula provides a physical interpretation of the K--R distribution
570: function in terms of a projection of the density operator into a combination
571: of squeezed states. Let us consider a linear superposition of
572: two infinitely squeezed coherent states
573: \begin{equation}
574: |\Phi\rangle={\cal N}^{\frac{1}{2}}
575: (|\alpha,-\Gamma\rangle+|\alpha,\Gamma\rangle). \label{sq}
576: \end{equation}
577: One can easily find that the real part of the K--R distribution, up to a
578: normalization factor, corresponds to a projection of the density operator on
579: the off-diagonal elements of the density matrix of the linear superposition
580: given by $|\Phi\rangle$. This connection with the
581: coherent squeezed states projection points on an important property of the K--R
582: function, that will be particularly useful in the phase space visualization of
583: various properties of squeezed states. The relation of the K--R distribution to
584: a projection on squeezed states, gives an operational meaning of such a phase
585: space distribution. The K--R phase space function of a given quantum state at
586: the phase space point $\alpha$ is just a projection of this state on
587: the off-diagonal elements of the density matrix of the
588: superposition of infinitely squeezed coherent states.
589:
590: \section{Quantum interference in phase space}\label{sec5}
591: All phase space functions from the Cohen class of distributions, Eq. (\ref{uog:wig}), are bilinear in
592: $\Psi$. This provides a transparent exhibition of quantum interference.
593: For a linear superposition of quantum states
594: \begin{equation*}
595: |\Psi\rangle=|\Psi_{1}\rangle+|\Psi_{2}\rangle ,
596: \end{equation*}
597: the corresponding phase space distribution function takes the form
598: \begin{equation*}
599: P=P_1+P_2+P_{int},
600: \end{equation*}
601: where $P_1$, $P_2$ correspond to distribution functions of states
602: $|\Psi_{1}\rangle$ and $|\Psi_{2}\rangle$, respectively, and $ P_{int}$
603: denotes the interference term. We shall study the properties of the
604: interference term for the superposition of two plane waves with momenta
605: $ p_{1} $ and $ p_{2}$ ($\hbar=1$):
606: \begin{equation}
607: \Psi \sim \exp(ip_{1}q)+\exp(ip_{2}q )\;. \label{sup}
608: \end{equation}
609: Based on these waves, we shall present a simple argument indicating the
610: differences between the quantum interference patterns exhibited with the help
611: of the Wigner function and the K--R function.
612:
613: The location of the interference terms of the Wigner function is well known.
614: The Wigner function for such a superposition is
615: \begin{eqnarray}
616: W(q,p) & \sim & \delta(p-p_{1})+\delta(p-p_{2}) \label{supwig}\\*
617: & + & 2\cos \left(q\negmedspace\vartriangle\negthickspace p\right)
618: \delta \left(p-\bar{p}\right),\nonumber
619: \end{eqnarray}
620: where $\vartriangle\negthickspace p=p_{2}-p_{1}$ and $\bar{p}=\frac{p_{1}+p_{2}}{2}$.
621: The characteristic interference term
622: $\cos \left(q\negmedspace\vartriangle\negthickspace p\right) $ is
623: isolated in phase space from the classical points described by the two momenta
624: $p_{1,2}$, and is located between these two incoherent terms at the mean
625: momentum $\bar{p}$.
626:
627: The generalized K--R function for such a superposition is given by
628: \begin{eqnarray}
629: & &K(q,p) \sim \delta(p-p_{1})+\delta(p-p_{2})\\
630: & +&
631: \delta\left(p-\bar{p}-\sigma\negthickspace\vartriangle\negthickspace p\right)
632: e^{i q \vartriangle\! p}+
633: \delta\left(p-\bar{p}+\sigma\negthickspace\vartriangle\negthickspace p\right)
634: e^{-i q \vartriangle\! p }, \nonumber
635: \end{eqnarray}
636: which leads to the following expression for the real part of the
637: generalized K--R distribution
638: \begin{eqnarray}
639: & &{\mathrm{Re}}[K(q,p)]=\delta(p-p_{1})+\delta(p-p_{2})
640: \label{suprrih}\\
641: & +&\cos\big(q\negmedspace\vartriangle\negthickspace p\big)
642: \Big[\delta\left(p-\bar{p}-\sigma\negthickspace\vartriangle\negthickspace p\right)+\delta\left(p-\bar{p}+\sigma\negthickspace\vartriangle\negthickspace p
643: \right)
644: \Big] .
645: \nonumber
646: \end{eqnarray}
647: We see that the interference term still oscillates like
648: $\cos\left(q\negmedspace\vartriangle\negthickspace p\right) $,
649: but the locations of these oscillations have moved to the points
650: different then in $W(q,p)$, Eq. (\ref{supwig}). Analyzing formula
651: (\ref{suprrih}) we find that parameter $\sigma$ shifts the
652: interference term along the momentum axis. For $\sigma=0$, which
653: corresponds to the Wigner function, interference term is located
654: at the mean momentum $\bar{p}$. For $\sigma=1$ (case of the K--R
655: distribution) these oscillations split and shift to the locations
656: defined by $p=p_1$ and $p=p_2$ in the phase space. Increase of
657: $\sigma$ beyond unity will move interference term further apart,
658: so that they not only no longer overlap each other but appear
659: outside the physical location of the state.
660:
661: The main difference between the quantum interference in the Wigner
662: representation and the K--R representation is the location of the
663: interference terms. This difference follows from the fact that the K--R
664: function depends locally on the phase space properties of the wave function,
665: while for the Wigner function this relation is nonlocal. This is why the
666: oscillations of the Wigner function occur at a position in phase space
667: which is different from the classical location.
668:
669:
670: It may appear, that the K--R function provides a less readable
671: representation of quantum interference, because the oscillating
672: terms cannot be isolated from the incoherent location. However,
673: the advantage of the K--R phase space representation is its local
674: relationship with the position and momentum wave functions.
675: The Wigner distribution function can have oscillations at
676: points where the wave function is vanishing. From the definition
677: of the K--R function we see that this phase space distribution has
678: to vanish when the wave functions $\Psi(q)$ and
679: $\widetilde{\Psi}(p)$ vanish in $q$ and $p$.
680:
681: \section{Linear superpositions in the K--R representation}\label{sec6}
682:
683: \subsection{Coherent states and Fock states}
684: As an example of the structures that can be obtained from Eq.
685: (\ref{k2}) we consider generalized K--R
686: distribution function
687: of a coherent state $|\alpha_0\rangle$.
688: By substituting $\hat{\rho}=|\alpha_0\rangle\langle \alpha_0| $ into
689: Eq. (\ref{k2}) after simple calculations we get the following
690: formula for generalized K--R distribution function of coherent
691: state
692: $|\alpha_0\rangle$:
693: \begin{eqnarray}
694: & &K_{\hat{\rho}}(\alpha,\sigma) =
695: \frac{2}{\pi\sqrt{1+\sigma\,^2}}\; \times \label{i:sigma}\\*
696: & & \exp\left(\frac{\sigma(\alpha^{\star}-
697: \alpha_0 ^{\star})^{2} }{1+\sigma\,^2}-
698: \frac{2|\alpha -\alpha_0|^{2}}{1+\sigma\,^2}
699: -\frac{\sigma(\alpha-\alpha_0)^{2} }{1+\sigma\,^2} \right).
700: \nonumber
701: \end{eqnarray}
702: For $\sigma=0$ it reduces to the well-known formula for the Wigner
703: distribution function of a coherent state. For other values of
704: $\sigma$ the generalized K--R distributions given by Eq.
705: (\ref{i:sigma}) also consist of a Gaussian bell shape centered
706: around a point ($\mathrm{Re}(\alpha_0)$, $\mathrm{Im}(\alpha_0)$)
707: but modified by a phase factor
708: \begin{equation}
709: e^{i\delta}=\exp\left[\frac{i 4 \sigma}{1+\sigma^2}
710: \left( \mathrm{Im}(\alpha)-\mathrm{Im}(\alpha_0)\right)\left(\mathrm{Re}(\alpha)-
711: \mathrm{Re}(\alpha_0)\right)
712: \right].
713: \nonumber
714: \label{factor1}
715: \end{equation}
716: The real part of this distribution has an oscillating term
717: $\cos\delta$, making the quasi-distribution non-positive.
718: For $\sigma\neq 0$, this oscillating term corresponds to the plane wave from
719: Eq. (\ref{prop2}) with the phase $\varphi(q,p)$ proportional to a
720: factor $\frac{4 \sigma}{1+\sigma^2}$. With increase of the value of
721: $\sigma$, these oscillations quickly become more and more rapid,
722: achieve maximum frequency for $\sigma=1$ and then slowly vanish.
723:
724:
725: Substituting $\alpha_0=0$ and $\sigma=1$ into the Eq. (\ref{i:sigma})
726: we obtain formula
727: that describes K--R distribution of the vacuum state $|0\rangle$:
728: \begin{equation}
729: K_{|0\rangle\langle 0|}(\alpha,1) =
730: \frac{\sqrt{2}}{\pi}\;
731: \exp\left(-
732: |\alpha|^{2}
733: -\frac{\alpha^{2}}{2}+\frac{(\alpha^{\star})^{2}}{2} \right).
734: \label{0_fock}
735: \end{equation}
736: \begin{figure}
737: \includegraphics[scale=.73]{af0.eps}
738: \caption{ The K--R distribution function for the vacuum state
739: $|0\rangle$.} \label{fock_zero}
740: \end{figure}
741: Figure \ref{fock_zero} shows the real part of the expression given
742: by Eq. (\ref{0_fock}). In accord with the previous description, it is
743: a Gaussian function in position and momentum, modulated by the
744: plane wave phase factor
745: $\cos[2\mathrm{Re}(\alpha)\mathrm{Im}(\alpha)]$.
746:
747: Formula for K--R distribution of the one photon Fock state
748: $|1\rangle$ is given by
749: \begin{eqnarray*}
750: K_{|1\rangle\langle 1|}(\alpha,1) =
751: \frac{\sqrt{2}(\alpha^2 - {\alpha^{\star}}^2) }{\pi}\times\\
752: \exp\left(-
753: |\alpha|^{2}
754: -\frac{\alpha^{2}}{2}+\frac{(\alpha^{\star})^{2}}{2}\right).
755: \end{eqnarray*}
756: The real part of this equation is shown in the Fig. \ref{fock_one}.
757: The difference between expressions for the K--R function of the
758: vacuum state and the one photon state is an additional multiplicative amplitude $i q p$ of the oscillating
759: phase factor. This amplitude comes from the product of the one photon wave
760: functions in momentum and position representations, respectively.
761: \begin{figure}
762: \includegraphics[scale=.73]{af1.eps}
763: \caption{ The K--R distribution function for the one photon Fock state. }
764: \label{fock_one}
765: \end{figure}
766:
767: \subsection{Superposition of coherent states}
768: Another simple but interesting and instructive example that we shall study is a
769: superposition of two coherent states $N(|\alpha_0\rangle +
770: |-\alpha_0\rangle)$. For this state, from Eq. (\ref{k2}) we derive:
771: %\begin{widetext}
772: \begin{eqnarray}
773: K(\alpha,\sigma)
774: =\frac{2 N}{\pi\sqrt{1+\sigma\,^2}}\;
775: \left[
776: e^{\frac{[\sigma(\alpha^{\star}-\alpha_0^{\star})^{2}-
777: \sigma(\alpha-\alpha_0)^{2}-2|\alpha -\alpha_0|^{2}]}
778: {1+\sigma\,^2}}
779: \right.\nonumber\\
780: +
781: e^{\frac{[\sigma(\alpha^{\star}+\alpha_0^{\star})^{2}-
782: \sigma(\alpha+\alpha_0)^{2}-2|\alpha +\alpha_0|^{2}]}
783: {1+\sigma\,^2}}
784: %\right.
785: \nonumber\\
786: +\left. e^{-2|\alpha_0|^2} \left(e^{\frac{[
787: \sigma(\alpha^{\star}+\alpha_0^{\star})^{2}-
788: \sigma(\alpha-\alpha_0)^{2}
789: -2
790: (\alpha^{\star}+\alpha_0^{\star})(\alpha -\alpha_0)]}{1+\sigma\,^2}}
791: \right.\right. \nonumber\\
792: +\left.\left.
793: e^{\frac{
794: [
795: \sigma(\alpha^{\star}-\alpha_0^{\star})^{2}-
796: \sigma(\alpha+\alpha_0)^{2}
797: -2
798: (\alpha^{\star}-\alpha_0^{\star})(\alpha +\alpha_0)]}{1+\sigma\,^2}}
799: \right) \right],
800: %\nonumber
801: \label{rihkot}
802: \end{eqnarray}
803: %\end{widetext}
804: where
805: \begin{equation*}
806: N=\Big(2+2e^{-2\mid\alpha_0\mid^2}\Big)^{-\frac{1}{2}}.
807: \end{equation*}
808:
809:
810:
811: Besides terms corresponding to two individual coherent states
812: there is an interference cross term. Its localization changes with
813: the change of parameter $\sigma$ in the same way as for the
814: superposition of two plane waves. This is illustrated in Fig.
815: \ref{s0}, where we have plotted the real part of Eq.
816: (\ref{rihkot}) for $\alpha_0=3$ and several values of parameter
817: $\sigma$. We present the real part of distributions rather then
818: the imaginary part (that does not contribute to marginals), or the
819: absolute value (in which no oscillations appear). In Fig.
820: \ref{s0c} we present the same functions using contour plots. In
821: this representation it is apparent how oscillating interference
822: term moves with the change of the value of the parameter $\sigma$.
823: As was the case of the superposition of plane waves, for
824: $\sigma=0$ interference term is located exactly between the terms
825: that do not correspond to the interference. With the increase of
826: the value of the parameter $\sigma$ interference terms split
827: away and move along ${\mathrm{Re}}(\alpha)$ axis, as they are
828: centered around phase space points $({\mathrm{Re}}(\alpha),
829: {\mathrm{Im}}(\alpha))=(\pm \sigma\alpha_0,0)$. Again, for
830: $\sigma=1$ oscillating interference terms overlap the terms
831: corresponding to the two individual coherent states.
832: As we have already mentioned, the K--R distribution function is
833: a special case of a distribution which is always equal to zero when
834: wave function is zero at certain point.
835:
836: Figure \ref{s0o} shows the same distribution functions when the
837: initial state is rotated in phase space. As an example we have
838: chosen previous initial state rotated by $\varphi=\frac{\pi}{4}$
839: in phase space, i.e. $\alpha_0=3\; e^{i\frac{\pi}{4}}$.
840: The Wigner function of
841: a ``rotated state'' is simply the rotated Wigner function. This
842: property is the basis for tomography of the Wigner function. By contrast, for other distributions (with parameter
843: $\sigma\neq 0$) this simple relation does not hold, and tomography
844: would not work.
845:
846: Let us consider
847: behavior of the terms corresponding to the individual coherent states in Fig. \ref{s0o}. Although the center of such a term is located exactly in this point
848: of phase space where coherent state is centered, closer
849: examination tell us that whole term is not ``rotated'' but rather
850: ``shifted'' by appropriate values parallel to the coordinate axes.
851: It is clearly seen that for $\sigma\neq 0$ generalized K--R
852: distribution functions single out $p$ and $q$ axes as compared to
853: the other directions in phase space. Interference terms are just
854: rotated but in the opposite direction.
855:
856: %%%%%%%%%%%%%%%%%%%%%%%%%
857: \begin{figure}
858: \noindent
859: a)
860: \includegraphics[scale=.7]{as0.eps}\\\noindent
861: b)\includegraphics[scale=.7]{as05.eps} \\\noindent
862: c)
863: \includegraphics[scale=.7]{as1.eps} \\
864: \caption{ Generalized K--R distribution
865: function of superposition of two coherent states
866: $N(|\alpha_0\rangle + |-\alpha_0\rangle)$. Here we have chosen
867: $\alpha_0=3$, and a) $\sigma=0$ (the Wigner distribution function),
868: b) $\sigma=0.5$, c) $\sigma=1$ (the K--R distribution function).}
869: \label{s0}
870: \end{figure}
871: %\newpage
872:
873: \begin{figure}
874: $\;\;$\\
875: $\;\;$\\
876: %\noindent
877: a)
878: \includegraphics[scale=.68]{as0c.eps}\\
879: \noindent
880: b)
881: \includegraphics[scale=.68]{as05c.eps}\\
882: $\;\;$\\
883: \noindent
884: c)
885: \includegraphics[scale=.68]{as1c.eps} \\
886: \caption{Contour plots of generalized K--R distribution
887: function of superposition of two coherent states
888: $N(|\alpha_0\rangle + |-\alpha_0\rangle)$.
889: Here we have chosen
890: $\alpha_0=3$, and a) $\sigma=0$ (the Wigner distribution function),
891: b) $\sigma=0.5$, c) $\sigma=1$ (the K--R distribution function).}
892: \label{s0c}
893: \end{figure}
894: %\newpage
895: \begin{figure}
896: \noindent
897: a)
898: \includegraphics[scale=0.68]{as0o.eps}\\
899: \noindent
900: b)
901: \includegraphics[scale=0.68]{as05o.eps}\\\noindent
902: c)
903: \includegraphics[scale=0.68]{as1o.eps}
904: \caption{Contour plots of generalized K--R distribution
905: function of superposition of two coherent states
906: $N(|\alpha_0\rangle + |-\alpha_0\rangle)$. Here we have chosen
907: $\alpha_0=3+i3$, and a) $\sigma=0$ (the Wigner distribution function),
908: b) $\sigma=0.5$, c) $\sigma=1$ (the K--R distribution function). }
909: \label{s0o}
910: \end{figure}
911: %\newpage
912: \begin{figure}
913: $\,$\\
914: $\xi=0.1$ \\
915: \includegraphics[scale=0.43]{asqz_01.eps}\\
916: $\,$\\
917: $\xi=0.5$\\
918: \includegraphics[scale=0.42]{asqz_05.eps}\\
919: $\,$\\
920: $\xi=0.7$\\
921: \includegraphics[scale=0.43]{asqz_07.eps}\\
922: $\,$\\
923: $\xi=0.9$\\
924: \includegraphics[scale=0.43]{asqz_09.eps}\\
925: $\,$\\$\,$\\
926: \caption{
927: The K--R distribution function of
928: superposition of two squeezed coherent states
929: ${\cal N}(|\xi\rangle + |-\xi\rangle)$. }
930: \label{sqz}
931: \end{figure}
932: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
933: \subsection{Superposition of coherent squeezed states}
934:
935: In the previous section we have connected the real part of the K--R function
936: with
937: the expectation value of off-diagonal elements of density matrix
938: corresponding to
939: superposition of two infinitely squeezed coherent states. It is interesting to
940: see how K--R distribution evolves if we change the squeezing parameter.
941:
942: %%%%
943: In Fig. \ref{sqz} we have plotted the real part of the K--R distribution function of
944: superposition of squeezed vacuum states
945: ${\cal N}(|\xi\rangle +
946: |-\xi\rangle)$ for several values of squeezing parameter $\xi$.
947: The K--R distribution apparently shows that an investigated state
948: consist of two perpendicularly squeezed states. Obviously, the
949: closer parameter $\xi$ gets to unity, the more significant are
950: effects of squeezing. Moreover, also basic properties of K--R
951: distribution are emphasized with the increase of the value of the
952: parameter $\xi$: the
953: oscillating $\cos[2\mathrm{Re}(\alpha)\mathrm{Im}(\alpha)]$-like
954: structure is more and more distinct. As we have mentioned before,
955: the K--R distribution single out $p$ a $q$ axes from other phase
956: space directions. Looking at the changes caused by the increase of
957: squeezing, one gets easily convinced of a close connection between the
958: K--R distribution and a superposition of infinitely squeezed
959: coherent states.
960:
961:
962:
963: \section{Summary}
964: We have presented a new class of phase space quasi-distribution
965: functions with correct momentum and position marginal properties,
966: that contains the Wigner distribution function and the K--R
967: distribution as special cases. We have shown how quantum
968: interference appears in phase space if such functions are used for
969: its investigation. In particular, we have focused on similarities
970: and differences between the Wigner function and the K--R
971: distribution.
972:
973: We have emphasized the most important properties of the K--R
974: distribution function: the fact that this function fully
975: characterizes the quantum state; that the K--R function
976: corresponds to the anti-standard ordering of $\hat{q}$ and
977: $\hat{p}$ operators; and that the real part of K--R distribution
978: is in natural way connected to coherent squeezed states.
979:
980: Most of the properties of the generalized K--R distribution
981: function we have presented using the formalism of coherent states
982: with the full analogy to the Glauber and Cahill s-ordered
983: quasi-distributions.
984:
985:
986: \section*{Acknowledgments}
987:
988: This work was partially supported by a KBN grant, {\em Splatanie
989: i interferencja atom\'ow i foton\'ow}, and the European Commission
990: through the Research Training Network QUEST.
991:
992:
993:
994: %\input{loglabs} %%% generates list of logical labels
995: %\input{bibcheck} %%% checks order of bibliography
996: \begin{thebibliography}{99}
997:
998: \bibitem{Wigner}
999: E. Wigner,
1000: Phys. Rev. {\bf 40}, 749 (1932)
1001:
1002: \bibitem{GlaCah}
1003: K. E. Cahill and R.J. Glauber,
1004: Phys. Rev. {\bf 177}, 1857, 1882
1005: (1969)
1006:
1007: \bibitem{Gla}
1008: R. J. Glauber,
1009: Phys. Rev. Lett. {\bf 10}, 84 (1963)
1010:
1011: \bibitem{Sud}
1012: E. C. G Sudarshan,
1013: Phys. Rev. Lett. {\bf 10}, 277 (1963)
1014:
1015: \bibitem{Husimi}
1016: K. Husimi, Proc. Phys. Math. Soc. Japan, {\bf 22}, 246 (1940)\\
1017: K. Takahashi, Suppl. Prog. Theor. Phys. {\bf 98}, 109 (1989)
1018:
1019:
1020: \bibitem{Kirkwood}
1021: J. G. Kirkwood,
1022: Phys. Rev. {\bf 44}, 31 (1933)
1023:
1024: \bibitem{Rihaczek}
1025: A. N. Rihaczek,
1026: IEEE Trans. Inf. Theory {\bf 14}, 369 (1968)
1027:
1028: \bibitem{Englert}
1029: B-G Englert, J. Phys. A, {\bf 22}, 625 (1989)
1030:
1031: \bibitem{Zak}
1032: J. Zak,
1033: Phys. Rev. A {\bf 45},
1034: 3540 (1992)
1035:
1036: \bibitem{O'Connell}
1037: R. F. O'Connell and E. P. Wigner, Phys. Lett. A vol 83 no 4 (1981), 145
1038:
1039: \bibitem{ff}
1040: W. P. Schleich, {\em Quantum Optics in Phase Space}, WILEY-VCH (2001).
1041:
1042:
1043: \bibitem{Cohen0}
1044: L. Cohen, Proceedings of the IEEE {\bf 77}, No. 7 (1989).
1045:
1046: \bibitem{Coh01}
1047: L. Cohen, {\em Time-Frequency Analysis}, PRENTICE HALL (1995).
1048:
1049:
1050: \bibitem{M-H}
1051: H. Margenau, R. N. Hill, Prog. Theor. Phys., {\bf 26}, 722 (1961)
1052:
1053:
1054: \bibitem{new}
1055: L. Praxmeyer, K. W\'odkiewicz, {\sl Hydrogen atom in the Kirkwood--Rihaczek
1056: representation} (in preparation)
1057:
1058:
1059: \bibitem{Cohen}
1060: L. Cohen,
1061: % {\em Generalized Phase-Space Distribution Functions},
1062: J. Math. Phys. {\bf 7}, 781 (1966)
1063:
1064:
1065: \bibitem{Cohen2}
1066: L. Cohen and Y.I. Zaparovanny,
1067: J. Math. Phys. {\bf 21}, 794 (1990)
1068:
1069:
1070:
1071: \bibitem{AgW}
1072: G. S. Agarwal and E. Wolf, Phys. Rev. D {\bf 2}, 2161, 2187, 2206
1073: (1970)
1074:
1075: \bibitem{squeeze}
1076: D. F. Walles, G. J. Milburn, {\em Quantum Optics}, SPRINGER--VERLAG
1077: (1994)
1078:
1079: \end{thebibliography}
1080:
1081:
1082: \end{document}
1083: