Debian Codegraph @

texmf-dist/doc/pdftex/tests/38-interword-space-at-font-switch/a1.tex

raw
1 \documentclass[12pt,reqno]{amsart}
2
3 \usepackage{amsmath,amsthm,amscd,amsfonts,amssymb,graphicx,color}
4
5 \usepackage[bookmarksnumbered,colorlinks,plainpages]{hyperref}
6 \hypersetup{colorlinks=true,linkcolor=red,anchorcolor=green,citecolor=cyan,urlcolor=red,filecolor=magenta,pdftoolbar=true}
7
8 \usepackage{mathrsfs}
9
10 % ----------------------------------------
11
12 \textheight 22.5truecm
13 \textwidth 14.5truecm
14 \setlength{\oddsidemargin}{0.35in}
15 \setlength{\evensidemargin}{0.35in}
16 \setlength{\topmargin}{-.5cm}
17
18 % ----------------------------------------
19
20 \newtheorem{theorem}{Theorem}[section]
21 \newtheorem{lemma}[theorem]{Lemma}
22 \newtheorem{proposition}[theorem]{Proposition}
23 \newtheorem{corollary}[theorem]{Corollary}
24 \theoremstyle{definition}
25 \newtheorem{definition}[theorem]{Definition}
26 \theoremstyle{remark}
27 \newtheorem{remark}[theorem]{Remark}
28 \newtheorem{notation}[theorem]{Notation}
29 \numberwithin{equation}{section}
30
31 % -------------------------------------------
32
33 \usepackage{microtype}
34 \begin{document}
35 \pdfcompresslevel=0
36 \pdfobjcompresslevel=0
37 \pdfinterwordspaceon
38
39 \setcounter{page}{1}
40
41 \title[A phase space localization operator in negative binomial states]{
42 A phase space localization operator in negative binomial states}
43
44 \author[Zouha\"{\i}r Mouayn, Soumia Touhami ]{
45 Zouha\"{\i}r Mouayn $^{1,2,3}$, Soumia Touhami $^4$}
46
47
48 % -----
49 \address{\footnotesize $^1$ Department of Mathematics, Faculty of Sciences and
50 Technics (M'Ghila), {\scriptsize Sultan Moulay Slimane University, B\'{e}ni Mellal, Morocco.}}
51 \address{\footnotesize $^2$ Institut des Hautes Études Scientifiques, Paris Saclay University
52 Le Bois-Marie, {\scriptsize 35 route de Chartres CS 40001 91893 Bures-sur-Yvette, France.}}
53 \address{\footnotesize $^3$ Institut Henri Poincaré - UAR 839 Sorbonne University, / CNRS, {\scriptsize 11 rue Pierre et Marie Curie
54 75231 Paris Cedex 05 France.}}
55 \address{\footnotesize $^4$ Department of Mathematics, KTH Royal Institute
56 of Technology, {\scriptsize Stockholm, Sweden.}}
57 \email{\textcolor[rgb]{0.00,0.00,0.84}{[email protected] $^{1}$; [email protected] $^{2}$}}
58
59
60 % -----------------------------------
61
62
63
64
65
66 \keywords{}
67
68 % -----------------------------------
69
70 \begin{abstract}
71 We are dealing with some spectral properties of a phase space localization
72 operator $P_{R}$ corresponding to the indicator function of a disk of radius
73 $R<1.$ The localization procedure is achieved with respect to a set of
74 negative binomial states (NBS) labeled by points of the complex unit disk\ $%
75 \mathbb{D}$ and depending on a parameter $B>%
76 %TCIMACRO{\U{bd}}%
77 %BeginExpansion
78 {\frac12}%
79 %EndExpansion
80 $ $.$ We derive a formula expressing $P_{R}$ as function of the
81 pseudo-harmonic oscillator whose potential function depends on $B$. The
82 phase space content outside the localization domain is estimated in terms of
83 the photon-counting probability distribution associated with the NBS. By
84 using the coherent states transform attached to NBS, we transfer the action
85 of the operator $P_{R}$ to a Bergman space\ $\mathcal{A}^{B}\left( \mathbb{D}%
86 \right) $ of analytic functions on\ $\mathbb{D}$ satisfying a growth
87 condition depending on $B$ and we explicitly give its integral kernel whose
88 limit\ as $R\rightarrow 1$ coincides with the reproducing kernel of \ $%
89 \mathcal{A}^{B}\left( \mathbb{D}\right) $. This leads to a natural
90 generalization of this Hilbert space with respect to the parameter $R$.
91
92 .
93 \end{abstract}
94
95 \maketitle
96
97 % -------------------------------------------
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127 \section{\ \textbf{Introduction}}
128
129 The problem of localization in time and frequency has always been of serious
130 concern in modern physics because one of the major issues in applications is
131 to analyze signals on different time-frequency domains and therefore to
132 concentrate and localize signals on these domains. To be able to represent
133 the frequency behavior of a signal locally in time, one has to consider the
134 so called time-frequency localization operators. A variety of methods have
135 been invented to construct such class of operators \cite{W}.
136 Coherent states (CS) are the natural tool in constructing phase space
137 localization operators \ and have been extensively encountered in
138 theoretical physics, in quantum mechanics and in many different areas of
139 mathematical physics. Precisely, CS provide a close connection between
140 classical and quantum formalisms so as to play a central role in the semi
141 classical analysis. In general, they may be defined as an overcomplete
142 family of normalized ket vectors $|\zeta \rangle $ which are labeled by
143 points $\zeta $ of a phase-space domain $X$, belonging to a Hilbert space $%
144 \mathcal{H}$ that corresponds to a specific quantum model and provide $%
145 \mathcal{H}$ with a resolution of its identity operator as
146 \begin{equation}
147 \label{r1}
148 \mathbf{1}_{\mathcal{H}}=\int_{X}|\zeta \rangle \langle \zeta |d\mu (\zeta ).
149 \end{equation}
150 with respect to a suitable integration measure $d\mu (\zeta )$ on $X$. These
151 states are constructed in different ways. For an overview of all aspects of
152 the theory of coherent states and their genesis, we refer to the \cite{V,TP}.
153
154 \smallskip
155
156 Equation $\eqref{r1} $ allows to implement a CS frame quantization \cite{G} of the set of parameters $\zeta \in X$ by associating to a
157 complex-valued function $\zeta \mapsto F(\zeta )$, satisfying appropriate
158 conditions, the following operator on $\mathcal{H}$ :
159 \begin{equation}
160 \label{r2}
161 F(\zeta )\mapsto P_{F}:=\int\limits_{X}|\zeta \rangle \langle \zeta |F(\zeta
162 )d\mu (\zeta ).
163 \end{equation}
164 If $F(\zeta )$ is semi-bounded real-valued function, the Friedrich extension
165 \cite{RS} allows us to define $P_{F}$ as a self-adjoint operator.
166 In particular, when $F=\chi _{\Omega }$ is the indicator function for some
167 domain $\Omega $ in the phase space $X$, the resulting operator $P_{\chi
168 _{\Omega }}$ is called a localization operator.
169
170 \smallskip \smallskip
171
172 By using the CS of the harmonic oscillator, Daubechies \cite{Daub}
173 has discussed the localization operator $P_{\chi _{\Omega }}$ with $\Omega
174 \subset \mathbb{C}$ being a disk of radius $\rho >0$ by giving its
175 eigenfunctions in terms of Hermite polynomials, and by expressing its
176 discrete eigenvalues $\left\{ \lambda _{k}^{\rho }\right\} $ in term of
177 incomplete Gamma functions. She also has established the asymptotic behavior
178 of these eigenvalues for varying $k=0,1,2,...,$ and $\rho >0$, and has given
179 an estimate for the phase-content outside the localization domain $\Omega $.
180
181 \smallskip
182
183 In this paper, we deal with similar questions for the pseudo-harmonic
184 oscillator
185
186 \begin{equation}
187 \label{r3}
188 H_{B}=\frac{1}{2}\left[ -\frac{d^{2}}{dx^{2}}+x^{2}+\frac{(2B-1)^{2}-\frac{1%
189 }{4}}{x^{2}}\right] +\left( 1-B\right) ,\text{ \ \ \ \ \ }2B>1
190 \end{equation}
191 acting on the Hilbert space $L^{2}\left( \mathbb{R}_{+}\right) $, whose
192 importance consists in the fact that it is a solvable model and being, in a
193 certain sense, an intermediate potential between the three dimensional
194 harmonic oscillator potential and other anharmonic potentials such as
195 Poschl-Teller or Morse potential \cite{PD1, PD2}. The $L^{2}$
196 eigenfunctions (number states) of $H_{B}$, which here are denoted by the ket
197 vectors $\left\vert \ell _{j}^{B}\right\rangle ,$ \ may be superposed to
198 perform a set of coherent states within the so-called \textit{Hilbertian
199 probabilistic scheme} (see \cite{G}for the general theory) by
200 choosing a set of analytic coefficients $C_{j}^{B}\left( z\right) $ on the
201 complex unit disk $\mathbb{D=}\left\{ z\in \mathbb{C},\left\vert
202 z\right\vert <1\right\} $ such that the associated photon-counting
203 statistics follows a negative probability distribution. Such a CS are known
204 as the negative binomials states (NBS) \cite{Mou1}.\ One interest on
205 them is that they intermediate between pure coherent states and pure thermal
206 states \cite{GJT}and reduce to Susskind-Glogower phases states for
207 a particular limit of the parameter \cite{FS} Beside, such
208 coefficients $C_{j}^{B}\left( z\right) $, turn out to be basis elements of
209 the weighted Bergman space, here denoted $\mathcal{A}^{B}\left( \mathbb{D}%
210 \right) $, of analytic functions $g$ on $\mathbb{D}$, satisfying the growth
211 condition $\int_{\mathbb{D}}|g(z)|^{2}(1-\bar{z}z)^{2B-2}d\eta (z)<+\infty $%
212 , where $d\eta $ denotes the Lebesgue measure on $\mathbb{D}$.
213
214 Our aim is, firstly, to show that these NBS which are labeled by points of
215 the disk $\mathbb{D}$ can be retreived from the affine CS via the Cayley
216 transform. We also link them to the Landau problem in the Poincar\'{e} upper
217 half-plane. This connection may be exploited to generalize the obtained
218 results to higher hyperbolic Landau levels. Secondly, we proceed by a
219 quantization method based on these NBS in order to construct a phase space
220 localization operator $P_{R}$ corresponding to the disk $D_{R}=\left\{ z\in
221 \mathbb{C},\left\vert z\right\vert <R\right\} $ with $R<1$, which stands for
222 the quantum counterpart of the classical observable defined as the indicator
223 function of the disk $D_{R}$. Precisely, we discuss some spectral properties
224 of the operator $P_{R}$ such as its eigenvalues and their associated
225 eigenfunctions in $L^{2}\left( \mathbb{R}_{+}\right) $. The expression of
226 these eigenvalues together with the discrete spectral resolution of $P_{R}$
227 amount to a formula expressing this operator as a function of the
228 Hamiltonian operator $H_{B}$ in $\eqref{r3} $. We also give an
229 estimate for the phase space content of $P_{R}$ outside the domain $D_{R}$
230 in terms of the photon-counting probability distribution associated with the
231 NBS. Moreover, this operator may be unitarly intertwined as $W\circ
232 P_{R}\circ W^{-1}$=$\widetilde{P}_{R}$ via the second Bargmann transform $W$
233 associated with the NBS. This allows us to obtain the integral kernel of $%
234 \widetilde{P}_{R}$ when acting on the space $\mathcal{A}^{B}\left( \mathbb{D}%
235 \right) $ by using calculations based on properties of some different
236 hypergeometric functions.
237
238 \smallskip
239
240 The paper is organized as follows. In section 2, we recall the affine
241 coherent states from which we derive the NBS.The connection with the Landau
242 problem on the Poincar\'{e} upper half-plane is also pointed out. Section 3
243 deals with the coherent states quantization method. In particular,
244 eigenvalues of the quantum counterpart with radial classical observables are
245 obtained.\ For the indicator function of the disk $D_{R}$ we also provide
246 these eigenvalues with a probabilistic interpretation and we discuss their
247 extensions to hyperbolic higher Landau levels. In section 4, we give an
248 estimate of the phase space content of the localization operator outside the
249 disk $D_{R}$ in terms of the photon-counting probablity distribution.
250 Section 5, we deal with the transfer of the localization operator to the
251 weigthed Bergman space $\mathcal{A}^{B}\left( \mathbb{D}\right) $ and to the
252 calculation of its integral kernel.
253
254
255
256
257
258
259
260
261
262
263 \section{ {\ \textbf{Negative binomial states and the $B$-weight Maass Laplacian}}}
264
265 \subsection{Affine coherent states}
266
267 We recall that the affine group is the set $\mathbf{G}=\mathbb{R}\times
268 \mathbb{R}^{+}$, endowed with group law $\left( x,y\right) \cdot \left(
269 x^{\prime },y^{\prime }\right) =\left( x+yx^{\prime },yy^{\prime }\right) $.
270 $\mathbf{G}$ is a locally compact group with the left Haar measure $d\nu
271 \left( x,y\right) =y^{-2}dxdy$. We shall consider one of the two
272 inequivalent infinite dimensional irreducible unitary representations of the
273 affine group $\mathbf{G}$, denoted $\pi _{+}$, realized on the Hilbert space
274 $\mathcal{H}:=$ $L^{2}\left( \mathbb{R}^{+},\xi ^{-1}d\xi \right) $ as
275 \begin{equation}
276 \label{r4}
277 \pi _{+}\left( x,y\right) \left[ \varphi \right] \left( \xi \right) :=e^{%
278 \frac{1}{2}ix\xi }\varphi \left( y\xi \right) ,\qquad \varphi \in \mathcal{\
279 H},\quad \xi >0\text{.}
280 \end{equation}
281 This representation is square integrable since it is easy to find a vector $%
282 \phi _{0}\in \mathcal{H}$ such that the function $\left( x,y\right) \mapsto
283 \left\langle \pi _{+}\left( x,y\right) \left[ \phi _{0}\right] ,\phi
284 _{0}\right\rangle _{\mathcal{H}}$ belongs to $L^{2}\left( \mathbf{G},d\nu
285 \right) $. This condition can also be expressed by saying that the
286 self-adjoint operator $\delta :\mathcal{H\rightarrow H}$ defined as $\delta %
287 \left[ \varphi \right] (\xi )=\xi ^{-\frac{1}{2}}\varphi \left( \xi \right) $
288 gives
289 \begin{equation}
290 \label{r5}
291 \int\limits_{\mathbf{G}}\left\langle \varphi _{1},\pi _{+}\left( x,y\right) %
292 \left[ \psi _{1}\right] \right\rangle \left\langle \pi _{+}\left( x,y\right) %
293 \left[ \varphi _{2}\right] ,\psi _{2}\right\rangle d\nu \left( x,y\right)
294 =\left\langle \varphi _{1},\varphi _{2}\right\rangle \left\langle \delta ^{%
295 \frac{1}{2}}\left[ \varphi _{1}\right] ,\delta ^{\frac{1}{2}}\left[ \varphi
296 _{2}\right] \right\rangle
297 \end{equation}
298 for all $\psi _{1},\psi _{2},\varphi _{1},\varphi _{2}\in \mathcal{H}$. The
299 operator $\delta $ is unbounded because $\mathbf{G}$ is not unimodular \cite{DMo}.
300
301 \smallskip
302
303 Keeping the condition $2B>1,$ we consider a set of CS labeled by elements $%
304 (x,y)\in \mathbf{G}$, which are obtained by acting, via the representation
305 operator $\pi _{+}\left( x,y\right) $, on the admissible vector
306 \begin{equation}
307 \label{r6}
308 \phi _{B}\left( \xi \right) :=\frac{1}{\sqrt{2B}}\xi ^{B}e^{-\frac{1}{2}\xi
309 },\text{ \ \ }\xi >0.
310 \end{equation}
311 Precisely,
312 \begin{equation}
313 \label{r7}
314 \left\vert \tau _{(x,y),B}\right\rangle :=\pi _{+}\left( x,y\right) \left[
315 \phi _{B}\right]
316 \end{equation}
317 and satisfy the resolution of the identity operator
318 \begin{equation}
319 \label{r8}
320 \mathbf{1}_{\mathcal{H}}=c_{B}\int\limits_{\mathbf{G}}d\mu \left( x,y\right)
321 \left\vert \tau _{(x,y),B}\right\rangle \left\langle \tau
322 _{(x,y),B}\right\vert
323 \end{equation}%
324 where $c_{B}:=2B-1$ and the Dirac's bra-ket notation $|\Phi \rangle \langle
325 \Phi |$ means the rank-one operator $\phi \longmapsto \langle \Phi ,\phi
326 \rangle _{\mathcal{H}}.\Phi $ with $\Phi ,\phi \in \mathcal{H}$. In the $\xi
327 $-coordinate, wavefunctions of CS defined by Eq. $\eqref{r7} $ read
328 \begin{equation}
329 \label{r9}
330 \left\langle \xi \right\vert \tau _{(x,y),B}\rangle =\frac{1}{\sqrt{2B}}%
331 \left( \xi y\right) ^{B}e^{-\frac{1}{2}\xi \left( y-ix\right) }\text{, }%
332 \qquad \xi >0,
333 \end{equation}
334 and are known as the affine CS \cite{AK}.
335
336
337
338
339 \subsection{Connection with the $B$-weight Maass Laplacian}
340
341 To describe the connection of CS $\eqref{r9} $ with the lowest
342 hyperbolic Landau level, we may first identify the affine group $\mathbf{G}$
343 with the Poincar\'{e} upper half-plane $\mathbb{H}^{2}=\left\{ x+iy,x\in
344 \mathbb{R},y>0\right\} $. Then, to these CS we may attach, as usual, the CS
345 transform $\mathcal{B}_{0}:\mathcal{H}\rightarrow L^{2}\left( \mathbb{H}%
346 ^{2},d\nu \right) $ defined by \cite{Mou2}:
347 \begin{equation}
348 \label{r10}
349 \mathcal{B}_{0}[\phi ]\left( x,y\right) =\sqrt{c_{B}}\int\limits_{0}^{+%
350 \infty }\overline{\left\langle \xi \right\vert \tau _{(x,y),B}\rangle }\phi
351 (\xi )\xi ^{-1}d\xi
352 \end{equation}
353 whose range is the eigenspace of the $B$-weight Maass Laplacian
354 \begin{equation}
355 \Delta _{B}=y^{2}\left( \partial _{x}^{2}+\partial _{y}^{2}\right)
356 -2iBy\partial _{x},
357 \end{equation}
358 associated with the eigenvalue
359 \begin{equation}
360 \label{r11}
361 \epsilon _{m}^{B}=(B-m)\left( 1-B+m\right) ,\quad m=0,1,...,\left\lfloor B-%
362 %TCIMACRO{\U{bd}}%
363 %BeginExpansion
364 {\frac12}%
365 %EndExpansion
366 \right\rfloor ,
367 \end{equation}
368 where $\left\lfloor a\right\rfloor $ denotes the greatest integer not
369 exceeding $a.$ We precisely have
370 \begin{equation}
371 \label{r12}
372 \mathcal{B}_{0}[\mathcal{H}]\equiv \left\{ f\in L^{2}\left( \mathbb{H}%
373 ^{2},d\nu \right) ,\;\Delta _{B}f=\epsilon _{0}^{B}f\right\} .
374 \end{equation}
375 The operator $\Delta _{B}$ also stands (in suitable units and up to an
376 additive constant) for the Schr\"{o}dinger operator describing the dynamics
377 of a charged particle moving on $\mathbb{H}^{2}$ under the action of a
378 magnetic field of strength proportional to $B.$ This is an elliptic densely
379 defined operator on the Hilbert space $L^{2}(\mathbb{H}^{2},d\nu )$, with a
380 unique self-adjoint realization also denoted by $\Delta _{B}$. Its spectrum
381 consists of two parts:\textit{\ }a continuous part $\left[
382 %TCIMACRO{\U{bc}}%
383 %BeginExpansion
384 {\frac14}%
385 %EndExpansion
386 \theta ,+\infty \right[ $, corresponding to scattering states and the finite
387 number of eigenvalues $\epsilon _{m}^{B}$ each one with infinite
388 degeneracy,\ called hyperbolic Landau levels . Finally, the reproducing
389 kernel of the Hilbert space $\mathcal{B}_{0}[\mathcal{H}]$ can be obtained
390 from the overlapping function $\langle \tau _{w,B},\tau _{\zeta ,B}\rangle _{%
391 \mathcal{H}}$ between two CS as
392 \begin{equation}
393 \label{r13}
394 K_{0}^{B}\left( w,\zeta \right) =\left( \frac{\left\vert w-\bar{\zeta}\right\vert ^{2}}{4\rm{Im}w\rm{Im}\zeta }\right) ^{-B}\left( \frac{\zeta-\bar{w}}{w-\bar{\zeta}}\right) ^{B}{},\text{ \ }w,\zeta \in \mathbb{H}^{2}.
395 \end{equation}
396
397
398
399 \subsection{Negative binomial states}
400
401 \smallskip We can write a version of these CS as states labeled by points $z$
402 of the unit disk $\mathbb{D}$ by using the inverse Cayley transform $%
403 \mathcal{C}^{-1}:\mathbb{D}\rightarrow \mathbf{G}$ given by
404 \begin{equation}
405 \label{r14}
406 \mathcal{C}^{-1}\left( z\right) =\left( -2\frac{\rm{Im}z}{\left\vert
407 1-z\right\vert ^{2}},\frac{1-\left\vert z\right\vert ^{2}}{\left\vert
408 1-z\right\vert ^{2}}\right) ,\text{ \ }z\in \mathbb{D}.
409 \end{equation}
410 Indeed, we may still define this version as states in $\mathcal{H}$ as%
411 \begin{equation}
412 \label{r15}
413 \kappa _{z,B}:=\left( \frac{1-\bar{z}}{1-z}\right) ^{B}\pi _{+}\left(
414 \mathcal{C}^{-1}\left( z\right) \right) \left[ \phi _{B}\right] .
415 \end{equation}
416 Direct calculations lead to their wave functions in the $\xi $-coordinate as
417 \begin{equation}
418 \label{r16}
419 \left\langle \xi \right\vert \kappa _{z,B}\rangle =\frac{1}{\sqrt{\Gamma
420 \left( 2B\right) }}\left( \frac{\left( 1-z\bar{z}\right) }{\left( 1-z\right)
421 ^{2}}\xi \right) ^{B}\exp \left( -\frac{1}{2}\left( \frac{1+z}{1-z}\right)
422 \xi \right) ,\qquad \xi >0.
423 \end{equation}
424 The latter ones may slightly be modified in order to perform them as vectors
425 of $L^{2}(\mathbb{R}_{+},d\xi )$ labeled by points of the disk $\mathbb{D}$
426 as
427 \begin{equation}
428 \label{r17}
429 \left\langle \xi \right\vert \widetilde{\kappa }_{z,B}\rangle :=\sqrt{\frac{2%
430 }{\xi }}\,\langle \xi ^{2}|\kappa _{z,B}\rangle
431 \end{equation}
432 These states obey the normalization condition $\langle \widetilde{\kappa }%
433 _{z,B},\widetilde{\kappa }_{z,B}\rangle _{L^{2}(\mathbb{R}_{+})}=1$ and
434 satisfy the resolution of the identity operator as
435 \begin{equation}
436 \label{r18}
437 1_{L^{2}(\mathbb{R}_{+})}=\int\limits_{\mathbb{D}}\left\vert \widetilde{%
438 \kappa }_{z,B}\right\rangle \left\langle \widetilde{\kappa }%
439 _{z,B}\right\vert d\eta _{B}(z),
440 \end{equation}
441 with respect to the measure
442 \begin{equation}
443 \label{r19}
444 d\eta _{B}(z):=\frac{(2B-1)}{\pi \left( 1-z\overline{z}\right) ^{2}}d\eta
445 (z).
446 \end{equation}
447
448 Now, as for the canonical CS of the harmonic oscillator, we seek for a
449 number states expansion of the CS $\left\vert \widetilde{\kappa }%
450 _{z,B}\right\rangle $ in terms of the analytic coefficients
451 \begin{equation}
452 \label{r20}
453 C_{j}^{B}(z):=(2B-1)^{1/2}\sqrt{\frac{\Gamma (2B+j)}{j!\Gamma (2B)}}z^{j},%
454 \text{ \ \ }j=0,1,2,...,
455 \end{equation}
456 which constitute an orthonormal basis of the weighted Bergman space $%
457 \mathcal{A}^{B}(\mathbb{D})$. For that, we start from the expression of the
458 CS, $\left\vert \widetilde{\kappa }_{z,B}\right\rangle $ as given by $\eqref{r17}-\eqref{r16}$ and we make use of the generating function
459 for the Laguerre polynomials (\cite{MaOb}, p.239):
460 \begin{equation}
461 \label{r21}
462 \sum_{j=0}^{+\infty }t^{j}L_{j}^{(\alpha )}(x)=\frac{1}{(1-t)^{\alpha +1}}%
463 \exp \left( \frac{-t}{1-t}x\right) ,\ \ \alpha >-1.
464 \end{equation}
465 We obtain, after some calculations, the series expansion
466 \begin{equation}
467 \label{r22}
468 \langle \xi \left\vert \widetilde{\kappa }_{z,B}\right\rangle =\left( \frac{%
469 2B-1}{(1-z\bar{z})^{2B}}\right) ^{-1/2}\sum_{j=0}^{+\infty }C_{j}^{B}(z)\ell
470 _{j}^{B}(\xi )
471 \end{equation}
472 where $\ell _{j}^{B}(\xi )$ are the Laguerre functions
473 \begin{equation}
474 \label{r23}
475 \ell _{j}^{B}(\xi ):=\left( \frac{2j!}{\Gamma (2B+j)}\right) ^{1/2}\xi ^{2B-%
476 \frac{1}{2}}e^{-\frac{1}{2}\xi ^{2}}L_{j}^{(2B-1)}(\xi ^{2}),\text{ \ \ }%
477 j=0,1,2,...,
478 \end{equation}
479 which are known to constitute a complete orthonormal system in $L^{2}(%
480 \mathbb{R}_{+},d\xi )$.
481
482 \smallskip
483
484 \begin{remark}
485
486 The CS $\eqref{r16} $ also coincide with those
487 constructed by Molanar \textit{et al} \cite{BMMB} for the Morse
488 potential \cite{MP} by an algebraic way based on supersymetry and
489 shape invariance properties where the shape parameter may be taken as our $%
490 B>0$. For this potential, they first were were introduced by Nieto \textit{%
491 et al} \cite{NMSL} as generalized minimal uncertainty
492 states.
493 \end{remark}
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528 \section{{\ \textbf{Quantization via $\ $NBS $\left\vert \widetilde{\protect\kappa }%
529 _{z,B}\right\rangle $}}}
530
531 The resolution of the identity $\eqref{r18}$ allows to implement a
532 CS or frame quantization \cite{TP} of the set of parameters $%
533 \mathbb{D}$ by associating to a function $\mathbb{D}\ni z\mapsto F(z,%
534 \overline{z})\in \mathbb{C}$ that satisfies appropriate conditions, the
535 following operator in $L^{2}(\mathbb{R}_{+},d\xi ):$
536 \begin{equation}
537 \label{r24}
538 F\mapsto \wp _{F}^{B}:=\int\limits_{\mathbb{D}}\left\vert \widetilde{\kappa }%
539 _{z,B}\right\rangle \left\langle \widetilde{\kappa }_{z,B}\right\vert F(z,%
540 \overline{z})\frac{(2B-1)}{\pi (1-z\bar{z})^{2}}d\eta (z).
541 \end{equation}
542 The Friedrich extension \cite{RS} allows to define $\wp _{F}^{B}$
543 as a self-adjoint operator if $F$ is a semi-bounded real-valued function.%
544 \newline
545
546 \bigskip \smallskip
547
548 In order to proceed with the quantization through the CS $\left\vert
549 \widetilde{\kappa }_{z,B}\right\rangle $ along the linear map $\eqref{r24} $, we may substitute the expression $\eqref{r22}$ into
550 the integral form in $\eqref{r24} .$ This gives the expression%
551 \begin{equation}
552 \label{r25}
553 \wp _{F}^{B}=\sum_{j,k=0}^{+\infty }\sqrt{\frac{\Gamma (2B+j)}{\pi j!\Gamma
554 (2B)}}\sqrt{\frac{\Gamma (2B+k)}{\pi k!\Gamma (2B)}}\left[
555 (2B-1)\int\limits_{\mathbb{D}}z^{k}\bar{z}^{j}(1-z\bar{z})^{2B-2}F(z,\bar{z}%
556 )d\eta (z)\right] \left\vert \ell _{j}^{B}\right\rangle \left\langle \ell
557 _{k}^{B}\right\vert
558 \end{equation}
559 which represents its disrete spectral resolution\textit{\ }%
560 \begin{equation}
561 \label{r26}
562 \wp _{F}^{B}=\sum_{j,k=0}^{+\infty }\left[ \gamma _{F}^{B}\right]
563 _{j,k}\left\vert \ell _{k}^{B}\rangle \langle \ell _{j}^{B}\right\vert
564 \end{equation}
565 where the matrix elements are (at least formally) given by\textit{\ }%
566 \begin{equation}
567 \label{r27}
568 \left[ \gamma _{F}^{B}\right] _{j,k}=\frac{1}{\pi \Gamma (2B-1)}\left( \frac{%
569 \Gamma (2B+j)\Gamma (2B+k)}{j!k!}\right) ^{1/2}\int\limits_{\mathbb{D}}\bar{z%
570 }^{j}z^{k}(1-z\bar{z})^{2B-2}F(z,\bar{z})d\eta (z)
571 \end{equation}
572 and\ $\left\{ \ell _{j}^{B}\right\} $ is the orthonormal basis of $L^{2}(%
573 \mathbb{R}_{+},d\xi ),$\ which is given in $\eqref{r23}.$
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602 \subsection{Radial classical observables}
603
604 For a radial weight function $F$, the above discrete spectral resolution of $%
605 \wp _{F}^{B}$ leads to more precise expressions for its eigenvalues. Indeed,
606 by setting $F(z,\bar{z})=F(r^{2})$, $r=\left\vert z\right\vert $ and using
607 polar coordinates in the expression $\eqref{r27} $ of matrix elements,
608 we get that
609 \begin{equation}
610 \label{r28}
611 \left[ \gamma _{F}^{B}\right] _{j,k}=\frac{1}{\pi \Gamma (2B-1)}\left( \frac{%
612 \Gamma (2B+j)\Gamma (2B+k)}{j!k!}\right) ^{1/2}\int\limits_{0}^{2\pi
613 }\int\limits_{0}^{1}r^{k}r^{j}e^{ik\theta }e^{-ij\theta
614 }(1-r^{2})^{2B-2}F(r^{2})rdrd\theta .
615 \end{equation}
616 By the fact that
617 \begin{equation}
618 \label{r29}
619 \int\limits_{0}^{2\pi }e^{i(k-j)\theta }d\theta =2\pi \delta _{k,j},\text{ \
620 }j,k=0,1,2,...,
621 \end{equation}
622 on can easily see that only the case $j=k$ produces a non zero matrix
623 element as
624 \begin{equation}
625 \label{r30}
626 \left[ \gamma _{F}^{B}\right] _{j,j}=\frac{\Gamma (2B+j)}{\Gamma
627 (2B-1)\Gamma (j+1)}\int_{0}^{1}\rho ^{j}(1-\rho )^{2B-2}F(\rho )d\rho .
628 \end{equation}
629 By writing the prefactor as $\left( \mathcal{B}(j+1,2B-1)\right) ^{-1}$, we
630 obtain the expression of the $\lambda _{j}^{B,F}$ as
631
632 \begin{equation}
633 \label{r31}
634 \lambda _{j}^{B,F}=\frac{1}{\mathcal{B}(j+1,2B-1)}\int\limits_{0}^{1}\rho
635 ^{j}(1-\rho )^{2B-2}F(\rho )d\rho ,
636 \end{equation}
637 where $\mathcal{B}(a,b)$ denotes the Beta function with $a,b>0$. The
638 operator $\wp _{F}^{B}$ has the following discrete spectral resolution with
639 respect to the orthonormal basis $\{\ell _{j}^{B}\}$ as
640 \begin{equation}
641 \label{r32}
642 \wp _{F}^{B}=\sum_{j=0}^{+\infty }\lambda _{j}^{B,F}\left\vert \ell
643 _{j}^{B}\right\rangle \left\langle \ell _{j}^{B}\right\vert ,
644 \end{equation}
645 and it's not difficult to check that
646 \begin{equation}
647 \label{r33}
648 \wp _{F}^{B}\left[ \ell _{j}^{B}\right] =\lambda _{j}^{B,F}\ell _{j}^{B}.
649 \end{equation}
650 Note that $\eqref{r33} $ means that the operators $\wp _{F}^{B}$ and $%
651 H_{B}$ have $\{\ell _{j}^{B}\}$ as a commun set of eigenfunctions.
652
653 \smallskip
654
655 In particular, we here consider the disk $D_{R}:=\{z\in \mathbb{C},\ |z|<R\}$
656 with $0<R<1$ and we choose as classical observable the indicator function of
657 this domain. By putting $F(r^{2})=1$ if $r<R$ and $F(r^{2})=0$ if $r\geq R$,
658 the formula $\eqref{r31} $ takes the form
659 \begin{equation}
660 \label{r34}
661 \lambda _{j}^{B,R}=\frac{1}{\mathcal{B}(j+1,2B-1)}\int\limits_{0}^{R^{2}}%
662 \rho ^{j}(1-\rho )^{2B-2}d\rho =\mathcal{I}_{R^{2}}(j+1,2B-1)
663 \end{equation}
664 where
665 \begin{equation}
666 \label{r35}
667 \mathcal{I}_{x}(a,b)=\frac{1}{\mathcal{B}(a,b)}\int%
668 \limits_{0}^{x}t^{a-1}(1-t)^{b-1}dt,\quad 0<x<1,\text{ }
669 \end{equation}
670 is the regularized incomplete Beta function. In view of $\eqref{r32}$, the discrete spectral resolution of $\wp _{\digamma }^{B}$
671 reads
672 \begin{equation}
673 \label{r36}
674 \wp _{R}^{B}=\sum_{j=0}^{+\infty }\mathcal{I}_{R^{2}}(j+1,2B-1)\left\vert
675 \ell _{j}^{B}\right\rangle \left\langle \ell _{j}^{B}\right\vert .
676 \end{equation}
677 By another side, the vector basis $\ell _{j}^{B}$ are eigenfunctions of $%
678 H_{B}$ while acting on $L^{2}\left( \mathbb{R}_{+},d\xi \right) $.
679 Precisely,
680 \begin{equation}
681 \label{r37}
682 H_{B}\left[ \ell _{j}^{B}\right] =(j+1)\ell _{j}^{B},\ j=0,1,2,...\text{ .}
683 \end{equation}
684 Therefore, we may write $\wp _{R}^{B}$ as a function of $H_{B}$ as
685 \begin{equation}
686 \label{r38}
687 \wp _{R}^{B}=\mathcal{I}_{R^{2}}(j+1,2B-1).
688 \end{equation}
689
690
691
692
693
694
695
696
697
698 \subsection{A probabilistic representation for eigenvalues $\protect\lambda %
699 _{j}^{B,R}$}
700
701 We note that the eigenvalues $\eqref{r34} $ may also be written as
702 \begin{equation}
703 \label{r39}
704 \lambda _{j}^{B,R}=\parallel C_{j}^{B}1_{\mathbb{D}_{R}}\parallel _{L^{2}(%
705 \mathbb{D},(1-z\bar{z})^{2B-2}d\eta )}^{2}=\int_{0}^{R^{2}}\mathfrak{g}%
706 _{j,B}(\rho )d\rho
707 \end{equation}
708 where
709 \begin{equation}
710 \label{r39}
711 \mathfrak{g}_{j,B}(\rho )=\frac{\Gamma (2B+j)}{\Gamma (2B-1)\Gamma (j+1)}%
712 (1-\rho )^{2B-2}\rho ^{j},\quad 0\leq \rho <1
713 \end{equation}
714 which turns out to be the identity function of the Beta distribution $%
715 \mathcal{Y}_{j,B}^{\left( 0\right) }\sim \mathcal{B}e(j+1,2B-1)$ whose
716 characteristic function is known to be given by the confluent hypergeometric
717 series as $u\mapsto {\ }_{1}F_{1}\left( j+1,2B+1;iu\right) $ with $i^{2}=-1.$%
718 Therefore,
719 \begin{equation}
720 \label{r40}
721 \lambda _{j}^{B,R}=\Pr (\mathcal{Y}_{j,B}^{\left( 0\right) }\leq R^{2})
722 \end{equation}
723 would provide us with a probabilistic representation of these eigenvalues.
724
725 \smallskip
726
727 For higher hyperbolic Landau levels$,$ the generalized form of the density
728 function $\eqref{r39}$ is given by
729 \begin{equation}
730 \label{r41}
731 \mathfrak{g}_{B,j}^{\left( m\right) }(\rho ):=\left( 2B-2m-1\right) \frac{%
732 \left( m\wedge j\right) !}{\left( m\vee j\right) !}\frac{\Gamma (2B-2m+m\vee
733 j)}{\Gamma (2B-2m+m\wedge j)}(1-\rho )^{2B-2m-2}\rho ^{\left\vert
734 m-j\right\vert }
735 \end{equation}
736 \begin{equation*}
737 \times \left( P_{m\wedge j}^{\left( \left\vert m-j\right\vert ,2\left(
738 B-m\right) -1\right) }\left( 1-2\rho \right) \right) ^{2}
739 \end{equation*}%
740 where $P_{k}^{\left( \alpha ,\beta \right) }$ $\left( .\right) $ $\ $is a
741 Jacobi polynomial \cite{MaOb} $\ $and $m=0,1,...,\left\lfloor B-%
742 %TCIMACRO{\U{bd}}%
743 %BeginExpansion
744 {\frac12}%
745 %EndExpansion
746 \right\rfloor .$ Let us denote by $\mathcal{Y}_{j,B}^{\left( m\right) }$ the
747 random variable having $\rho \mapsto \mathfrak{g}_{j,B}^{\left( m\right)
748 }(\rho )$ as its density, then
749 \begin{equation}
750 \label{r42}
751 \lambda _{j}^{B,R,m}:=\Pr \left( \mathcal{Y}_{j,B}^{\left( m\right) }\leq
752 R^{2}\right) =\int\limits_{0}^{R^{2}}\mathfrak{g}_{j,B}^{\left( m\right)
753 }(\rho )d\rho
754 \end{equation}
755 would provide us with the probabilistic representation of eigenvalues $%
756 \lambda _{j}^{B,R,m}$ of the restricted operator $\mathfrak{K}_{B,m}\mid
757 _{D_{R}}$to the disk $D_{R},$ where $\mathfrak{K}_{B,m}$ is the projection
758 operator onto the eigenspace
759 \begin{equation}
760 \label{r43}
761 \mathcal{E}_{B,m}\left( \mathbb{D}\right) =\left\{ f\in L^{2}\left( \mathbb{D%
762 },\left( 1-z\overline{z}\right) ^{2B-2}d\eta \left( z\right) \right) ,%
763 \widetilde{\Delta }_{B}f=\sigma _{B,m}f\right\}
764 \end{equation}
765 of the $B$-weight Maass Laplacian
766 \begin{equation}
767 \label{r44}
768 \widetilde{\Delta }_{B}=-4\left( 1-z\overline{z}\right) \left( \left( 1-z%
769 \overline{z}\right) \frac{\partial ^{2}}{\partial z\partial \overline{z}}-2B%
770 \overline{z}\frac{\partial }{\partial \overline{z}}\right) ,
771 \end{equation}
772 associated with the hyperbolic Landau level%
773 \begin{equation}
774 \label{r45}
775 \sigma _{B,m}=4m\left( 2B-m-1\right) ,\text{ \ \ }m=0,1,...,\left\lfloor B-%
776 %TCIMACRO{\U{bd}}%
777 %BeginExpansion
778 {\frac12}%
779 %EndExpansion
780 \right\rfloor .
781 \end{equation}
782 So that the integral kernel of operator $\mathfrak{K}_{B,m}$ is the
783 reproducing kernel
784 \begin{equation}
785 \label{r46}
786 K_{B,m}\left( z,w\right) =\pi \left( 2B-2m-1\right) \left( 1-z\overline{w}%
787 \right) ^{-2B}\left( \frac{\left\vert 1-z\overline{w}\right\vert ^{2}}{%
788 \left( 1-z\overline{z}\right) \left( 1-w\overline{w}\right) }\right) ^{m}
789 \end{equation}
790 \begin{equation*}
791 \times P_{m}^{\left( 0,2\left( B-m\right) -1\right) }\left( 2\frac{\left( 1-z%
792 \overline{z}\right) \left( 1-w\overline{w}\right) }{\left\vert 1-z\overline{w%
793 }\right\vert ^{2}}-1\right)
794 \end{equation*}
795 of the eigenspace $\eqref{r46}.$ Note that for $2B>1$ and $m=0$ this
796 eigenspace reduces to the weighted Bergman space of analytic functions $g$
797 on $\mathbb{D}$, satisfying the growth condition $\int_{\mathbb{D}%
798 }|g(z)|^{2}(1-\bar{z}z)^{2B-2}d\eta (z)<+\infty $. In other words, $\mathcal{%
799 E}_{B,0}\left( \mathbb{D}\right) \equiv \mathcal{A}^{B}\left( \mathbb{D}%
800 \right) .$ This remark makes possible to extend our analysis to the above
801 higher hyperbolic Landau levels $\sigma _{B,m}.$ In this respect the
802 connection with the results in \cite{CDM} may be useful.
803
804
805
806
807
808
809
810
811
812
813
814
815 \section{{\ \textbf{Phase space content of $\wp _{R}^{B}$ outside $D_{R}$}}}
816
817 Note that the phase space cutoff by the operator $\wp _{R}^{B}$ is not sharp
818 in the sense that it will have some phase space content outside the
819 localization domain $D_{R}$. This is illustrated by the fact that at least
820 for some coherence point $z_{0}\in \mathbb{D}\setminus D_{R}$ we have
821 \begin{equation}
822 \label{r47}
823 \langle \widetilde{\kappa }_{z_{0},B}\left\vert \wp _{R}^{B}[f]\right\rangle
824 _{L^{2}(\mathbb{R}_{+})}\neq 0,\quad f\in L^{2}(\mathbb{R}_{+}).
825 \end{equation}
826 Precisely, we have the following estimate involving the photon counting
827 statistics which obey the negative binomial probability distribution $%
828 \mathcal{X}$ $\sim \mathcal{NB}(2B,z_{0}\bar{z}_{0})$\textit{\ }with\textit{%
829 \ }parameters $2B$\ and $z_{0}\bar{z}_{0}$
830 \begin{equation}
831 \label{r48}
832 \left\vert \langle \widetilde{\kappa }_{z_{0},B}\left\vert \wp
833 _{R}^{B}[f]\right\rangle _{L^{2}\left( \mathbb{R}_{+}\right) }\right\vert
834 \leq \sqrt{\mathbb{E}\left( \left( \mathcal{I}_{R^{2}}\left( \mathcal{X+}%
835 1,2B-1\right) \right) ^{2}\right) }\Vert f\Vert _{L^{2}\left( \mathbb{R}%
836 _{+}\right) } .
837 \end{equation}
838 To prove $\eqref{r48}$ we start by replacing in the scalar product
839 $\eqref{r47}$ the operator $\wp _{R}^{B}$ by its discrete spectral
840 resolution $\eqref{r36}$ as
841 \begin{equation}
842 \label{r49}
843 \left\langle \widetilde{\kappa }_{z_{0},B},\left( \sum_{j=0}^{+\infty
844 }\lambda _{j}^{B,R}\left\vert \ell _{j}^{B}\right\rangle \left\langle \ell
845 _{j}^{B}\right\vert \right) [f]\right\rangle _{L^{2}(\mathbb{R}%
846 _{+})}=\sum_{j=0}^{+\infty }\lambda _{j}^{B,R}\left\langle \widetilde{\kappa
847 }_{z_{0},B}|\ell _{j}^{B}\right\rangle _{L^{2}\left( \mathbb{R}_{+}\right)
848 }\left\langle \ell _{j}^{B}|f\right\rangle _{L^{2}\left( \mathbb{R}%
849 _{+}\right) }.
850 \end{equation}
851 Recalling the number states expansion $\eqref{r22} $ of $\left\vert
852 \widetilde{\kappa }_{z,B}\right\rangle $, we may write
853 \begin{equation}
854 \label{r50}
855 \left\langle \widetilde{\kappa }_{z_{0},B}|\ell _{j}^{B}\right\rangle
856 _{L^{2}\left( \mathbb{R}_{+}\right) }=(1-z_{0}\bar{z}_{0})^{B}\sqrt{\frac{%
857 \Gamma (2B+j)}{j!\Gamma (2B)}}z_{0}^{j}.
858 \end{equation}
859 Therefore $\eqref{r49} $ reads
860 \begin{equation}
861 \label{r51}
862 \langle \widetilde{\kappa }_{z_{0},B}\left\vert \wp _{R}^{B}[f]\right\rangle
863 _{L^{2}\left( \mathbb{R}_{+}\right) }=(1-z_{0}\bar{z_{0}})^{B}\sum_{j=0}^{+%
864 \infty }\lambda _{j}^{B,R}\left( \frac{\Gamma (2B+j)}{j!\Gamma (2B)}\right)
865 ^{1/2}\left\langle \ell _{j}^{B}|f\right\rangle _{L^{2}\left( \mathbb{R}%
866 _{+}\right) }z_{0}^{j}.
867 \end{equation}
868 \begin{equation}
869 \label{r52}
870 =(1-z_{0}\bar{z_{0}})^{B}\left\langle \sum_{j=0}^{+\infty }\lambda
871 _{j}^{B,R}\left( \frac{\Gamma (2B+j)}{j!\Gamma (2B)}\right)
872 ^{1/2}z_{0}^{j}\ell _{j}^{B}|f\right\rangle _{L^{2}(\mathbb{R}_{+})}.
873 \end{equation}
874 Setting
875 \begin{equation}
876 \label{r53}
877 \vartheta :=\sum_{j=0}^{+\infty }\lambda _{j}^{B,R}\left( \frac{\Gamma (2B+j)%
878 }{j!\Gamma (2B)}\right) ^{1/2}z_{0}^{j}\ell _{j}^{B},
879 \end{equation}
880 and using the Cauchy-Bunyakovsky-Schwarz inequality
881 \begin{equation}
882 \label{r54}
883 \left\vert \langle \widetilde{\kappa }_{z_{0},B}|\wp _{R}^{B}[f]\rangle
884 _{L^{2}\left( \mathbb{R}_{+}\right) }\right\vert \leq |1-z_{0}\bar{z_{0}}%
885 |^{B}\Vert \vartheta \Vert _{L^{2}(\mathbb{R}_{+})}\left\Vert f\right\Vert
886 _{L^{2}\left( \mathbb{R}_{+}\right) }.
887 \end{equation}
888 Now, the $L^{2}$ norm of $\vartheta $ may be written
889 \begin{equation}
890 \label{r54}
891 |1-z_{0}\bar{z_{0}}|^{B}\Vert \vartheta \Vert _{L^{2}(\mathbb{R}%
892 _{+})}=\left( \sum_{j=0}^{+\infty }\left( \lambda _{j}^{B,R}\right) ^{2}%
893 \left[ \frac{\Gamma (2B+j)}{j!\Gamma (2B)}(z_{0}\bar{z_{0}})^{j}(1-z_{0}\bar{%
894 z_{0}})^{2B}\right] \right) ^{1/2}.
895 \end{equation}
896 By recognizing in $\eqref{r54} $ the probability distribution of $%
897 \mathcal{X}\sim \mathcal{N}\mathcal{B}(2B,z_{0}\overline{z_{0}})$ :
898 \begin{equation}
899 \label{r55}
900 \Pr \left( \mathcal{X}=j\right) =\frac{\Gamma (2B+j)}{j!\Gamma (2B)}\left(
901 z_{0}\overline{z_{0}}\right) ^{j}(1-z_{0}\overline{z_{0}})^{2B},\ \
902 j=0,1,2,\cdots ,
903 \end{equation}
904 and using the expression of the eigenvalues $\lambda _{j}^{B,R}$, Eq.$\eqref{r54} $ takes the form
905 \begin{equation}
906 \label{r56}
907 |1-z_{0}\bar{z_{0}}|^{B}\Vert \vartheta \Vert _{L^{2}(\mathbb{R}%
908 _{+})}=\left( \sum_{j=0}^{+\infty }\left( \mathcal{I}_{R^{2}}(j+1,2B-1)%
909 \right) ^{2}\Pr \left( \mathcal{X}=j\right) \right) ^{1/2}.
910 \end{equation}
911 Finally, the right hand side of $\eqref{r56} $ can be viewed as the
912 square root of an expectation value as $\mathbb{E}\left( \left( \mathcal{I}%
913 _{R^{2}}(\mathcal{X}+1,2B-1)\right) ^{2}\right) $.
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935 \section{\textbf{Integral kernel of $\widetilde{\wp }_{R}^{B}$ \ on the Bergman
936 space $\mathcal{A}^{B}(\mathbb{D})$}}
937
938 The CS transform associated with $\left\vert \widetilde{\kappa }%
939 _{z,B}\right\rangle $ is the isometric isomorphism $W_{B}:L^{2}(\mathbb{R}%
940 _{+})\rightarrow \mathcal{A}^{B}(\mathbb{D})$ defined by $\varphi \mapsto
941 W_{B}[\varphi ](z):=(1-z\bar{z})^{-B}\langle \varphi \left\vert \widetilde{%
942 \kappa }_{z,B}\right\rangle _{L^{2}\left( \mathbb{R}_{+}\right) }$, which
943 may also be called the second Bargmann transform \cite{IWGM}.
944 Explicitly,
945 \begin{equation}
946 \label{r5.1}
947 W_{B}[f](z)=\sqrt{\frac{2}{\Gamma (2B)}}(1-z)^{-2B}\int_{0}^{+\infty }\xi
948 ^{2B-\frac{1}{2}}\exp \left( -\frac{1}{2}\left( \frac{1+z}{1-z}\right) \xi
949 ^{2}\right) d\xi ,\ z\in \mathbb{D}.
950 \end{equation}
951 Now, to transfer the operator $\wp _{R}^{B}$ to be acting on functions $f$ $%
952 \in $ $\mathcal{A}^{B}(\mathbb{D}),$ we use the relation $W_{B}\circ \wp
953 _{R}^{B}\circ W_{B}^{-1}\equiv \widetilde{\wp }_{R}^{B}$. \ We, precisely,
954 obtain that
955 \begin{equation}
956 \label{r5.2}
957 \widetilde{\wp }_{R}^{B}[f](w)=\int\limits_{\mathbb{D}}P_{R}^{B}(z,w)f(z)(1-z%
958 \bar{z})^{2B-2}d\eta (z)
959 \end{equation}
960 where the integral kernel is given by
961 \begin{equation}
962 \label{r5.3}
963 P_{R}^{B}(z,w)=\frac{(2B-1)^{2}R}{(1-R\bar{z}w)^{2B}}F_{1}\left(
964 2-2B,1-2B,2B,2,R,\frac{R-R\bar{z}w}{1-R\bar{z}w}\right)
965 \end{equation}
966 In particular, at the limit $R\rightarrow 1$,
967 \begin{equation}
968 \label{r5.4}
969 \lim_{R\rightarrow 1}P_{R}^{B}(z,w)=\frac{(2B-1)}{\left( 1-\bar{z}w\right)
970 ^{2B}},
971 \end{equation}
972 which is the reproducing kernel of the Bergman space $\mathcal{A}^{B}(%
973 \mathbb{D}).$ Here, $F_{1}$ denotes the Appel hypergeometric double series%
974 \begin{equation*}
975 F_{1}\left( \alpha ,\beta ,\gamma ;\omega ;u,v\right)
976 =\sum\limits_{j,k=0}^{\infty }\frac{\left( \alpha \right) _{j+k}\left( \beta
977 \right) _{j}\left( \gamma \right) _{k}}{j!k!\left( \omega \right) _{j+k}}%
978 u^{j}v^{k},\left\vert u\right\vert <1,\left\vert v\right\vert <1.
979 \end{equation*}
980
981 \bigskip \smallskip
982
983 To prove $\eqref{r5.2} ,$ let us take $f\in \mathcal{A}^{B}(\mathbb{D})$
984 and apply the inverse of $W_{B}$ as
985 \begin{equation}
986 \label{r5.5}
987 W_{B}^{-1}[f](\xi )=\int\limits_{\mathbb{D}}f(z)\left\langle \xi \right\vert
988 \widetilde{\kappa }_{z,B}\rangle (1-z\bar{z})^{-B}d\eta _{B}(z),\text{ }\xi
989 >0.
990 \end{equation}
991 Next, we proceed by the action of $\wp _{R}^{B}$ on $W_{B}^{-1}[f],$ which
992 successively gives
993
994 \begin{equation}
995 \label{r5.6}
996 \wp _{R}^{B}\left[ W_{B}^{-1}[f]\right] (y)=\sum_{k=0}^{+\infty }\lambda
997 _{k}^{B,R}\left\langle \ell _{k}^{B}|\int\limits_{\mathbb{D}}f(z)|\widetilde{%
998 \kappa }_{z,B}\rangle (1-z\bar{z})^{-B}d\eta _{B}(z)\right\rangle
999 \left\langle y|\ell _{k}^{B}\right\rangle
1000 \end{equation}
1001
1002 \begin{equation}
1003 \label{r5.7}
1004 =\sum_{k=0}^{+\infty }\lambda _{k}^{B,R}\left( \int_{\mathbb{R}_{+}}%
1005 \overline{\ell _{k}^{B}(\xi )}\left( \int_{\mathbb{D}}f(z)\left\langle
1006 x\right\vert \widetilde{\kappa }_{z,B}\rangle (1-z\bar{z})^{-B}d\eta
1007 _{B}(z)\right) d\xi \right) \ell _{k}^{B}(y)
1008 \end{equation}
1009
1010 \begin{equation}
1011 \label{r5.8}
1012 =\sum_{k=0}^{+\infty }\lambda _{k}^{B,R}\int_{\mathbb{D}}f(z)\left( (1-z\bar{%
1013 z})^{-B}\overline{\int_{\mathbb{R}_{+}}\ell _{k}^{B}(\xi )\overline{%
1014 \left\langle \xi \right\vert \widetilde{\kappa }_{z,B}\rangle }}dx\right)
1015 d\eta _{B}\left( z\right) \ell _{k}^{B}(y).
1016 \end{equation}
1017 Note that the last equation may be rewritten as
1018 \begin{equation}
1019 \label{r5.9}
1020 \wp _{R}^{B}\left[ W_{B}^{-1}[f]\right] (y)=\sum_{k=0}^{+\infty }\lambda
1021 _{k}^{B,R}\left[ \int_{\mathbb{D}}f(z)\overline{W_{B}[\ell _{k}^{B}](z)}%
1022 d\eta _{B}(z)\right] \ell _{k}^{B}(y).
1023 \end{equation}
1024 We again apply $W_{B}$ to $\eqref{r5.9}:$
1025 \begin{equation}
1026 \label{r5.10}
1027 W_{B}\left[ \wp _{R}^{B}\left[ W_{B}^{-1}[f]\right] \right]
1028 (w)=\sum_{k=0}^{+\infty }\lambda _{k}^{B,R}\left[ \int_{\mathbb{D}}f(z)%
1029 \overline{W_{B}[\ell _{k}^{B}](z)}d\eta _{B}(z)\right] W_{B}[\ell
1030 _{k}^{B}](w).
1031 \end{equation}
1032 Since $W_{B}[\ell _{k}^{B}](w)=C_{k}^{B}(w),$ then we obtain
1033 \begin{equation}
1034 \label{r5.11}
1035 \widetilde{\wp }_{R}^{B}[f](w)=\int_{\mathbb{D}}\left[ \sum_{k=0}^{+\infty
1036 }\lambda _{k}^{B,R}\overline{C_{k}^{B}(z)}C_{k}^{B}(w)\right] d\eta _{B}(z).
1037 \end{equation}
1038 Hence, the kernel integral function is given by
1039 \begin{equation}
1040 \label{r5.12}
1041 P_{R}^{B}(z,w)=\sum_{k=0}^{+\infty }\lambda _{k}^{B,R}\overline{C_{k}^{B}(z)}%
1042 C_{k}^{B}(w).
1043 \end{equation}
1044 To write this kernel in a closed form, we recall $\eqref{r34} $ and $\eqref{r20} $, then $\eqref{r5.12} $ becomes%
1045 \begin{equation}
1046 \label{r5.13}
1047 P_{R}^{B}(z,w)=(2B-1)\sum_{k=0}^{+\infty }\left[ \frac{1}{B(k+1,2B-1)}%
1048 \int_{0}^{R}\rho ^{k}(1-\rho )^{2B-2}d\rho \right] \frac{\Gamma (2B+k)}{%
1049 k!\Gamma (2B)}\bar{z}^{k}w^{k}
1050 \end{equation}
1051 \begin{equation}
1052 \label{r5.14}
1053 =(2B-1)^{2}\sum_{k=0}^{+\infty }\frac{\Gamma (2B+k)\Gamma (2B+k)}{\Gamma
1054 (2B)\Gamma (2B)}\frac{1}{k!}\frac{(\bar{z}w)^{k}}{k!}\left(
1055 \int_{0}^{R}t^{k}(1-t)^{2B-2}dt\right)
1056 \end{equation}
1057 \begin{equation}
1058 \label{r5.15}
1059 =(2B-1)^{2}\int_{0}^{R}(1-t)^{2B-2}\left( \sum_{k=0}^{+\infty }\frac{%
1060 (2B)_{k}(2B)_{k}}{(1)_{k}}\frac{(t\bar{z}w)^{k}}{k!}\right) dt
1061 \end{equation}
1062 The sum inside the integral can be presented as the Gauss hypergeometric
1063 function ${}_{2}F_{1}$ as
1064 \begin{equation}
1065 \label{r5.16}
1066 P_{R}^{B}(z,w)=(2B-1)^{2}\int_{0}^{R}(1-t)^{2B-2}{}_{2}F_{1}(2B,2B,1,t\bar{z}%
1067 w)dt.
1068 \end{equation}
1069 By setting $2B=\alpha $, $\bar{z}w=\omega $ and
1070 \begin{equation}
1071 \label{r5.17}
1072 I_{R}=\int_{0}^{R}{}_{2}F_{1}(\alpha ,\alpha ,1;\omega t)(1-t)^{\alpha -2}dt.
1073 \end{equation}
1074 By making use of the formula \cite[p.316]{PBM}
1075 :
1076 \begin{equation}
1077 \label{r5.18}
1078 \int_{0}^{y}x^{c-1}(y-x)^{\beta -1}(1-xz)^{-\tau }{}_{2}F_{1}(a,b,c;wx)dx=%
1079 \mathcal{B}(c,\beta )y^{c+\beta -1}(1-yz)^{-\tau }F_{3}(\tau ,a,\beta
1080 ,b,c+\beta ,\frac{yz}{yz-1};wy),
1081 \end{equation}
1082 $y,Re(c),Re(\beta )>0;|arg(1-wy)|,|arg(1-z)|<\pi $, for parameters $%
1083 y=R,x=t,c=1,\beta =1,z=1,\tau =2-2B,a=2B,b=2B$, the integral $\eqref{r5.17} $ takes the form
1084 \begin{equation}
1085 \label{r5.19}
1086 I_{R}=R(1-R)^{\alpha -2}F_{3}(2-\alpha ,\alpha ,1,\alpha ,2;\frac{R}{R-1}%
1087 ,\omega R)
1088 \end{equation}
1089 Next, we may apply the transformation \cite[p.450]{PBM}:
1090 \begin{equation}
1091 \label{r5.20}
1092 F_{3}\left( a,a^{\prime },b,b^{\prime };a+a^{\prime },w,z\right)
1093 =(1-z)^{-b^{\prime }}F_{1}\left( a,b,b^{\prime };a+a^{\prime };w,\frac{z}{z-1%
1094 }\right)
1095 \end{equation}
1096 to rewrite the $F_{3}$ hypergeometric function in $\eqref{r5.19} $ as
1097 \begin{equation}
1098 \label{r5.21}
1099 F_{3}\left( 2-\alpha ,\alpha ,1,\alpha ,\frac{R}{R-1},\omega R\right)
1100 =(1-\omega R)^{-\alpha }F_{1}\left( 2-\alpha ,1,\alpha ,2,\frac{R}{R-1},%
1101 \frac{R\omega }{R\omega -1}\right) .
1102 \end{equation}%
1103 Therefore, Eq.$\eqref{r5.19} $ reads
1104 \begin{equation}
1105 \label{r5.22}
1106 I_{R}=R(1-R)^{\alpha -2}(1-\omega R)^{-\alpha }F_{1}\left( 2-\alpha
1107 ,1,\alpha ,2,\frac{R}{R-1},\frac{R\omega }{R\omega -1}\right) .
1108 \end{equation}
1109 By applying the transformation
1110 \begin{equation}
1111 \label{r5.23}
1112 F_{1}\left( a,b_{1},b_{2},c;X;Y\right)
1113 =(1-X)^{-b_{1}}(1-Y)^{-b_{2}}F_{1}\left( c-a,b_{1},b_{2};c,\frac{X}{X-1},%
1114 \frac{Y}{Y-1}\right)
1115 \end{equation}
1116 we may reduce $I_{R}$ as
1117 \begin{equation}
1118 \label{r5.24}
1119 I_{R}=R(1-R)^{\alpha -1}F_{1}\left( \alpha ,1,\alpha ,2;R,R\omega \right) .
1120 \end{equation}
1121 Next, by using the symmetry relation
1122 \begin{equation}
1123 \label{r5.25}
1124 F_{1}\left( a,b,b^{\prime };c,z,u\right) =F_{1}\left( a,b^{\prime
1125 },b;c,u,z\right)
1126 \end{equation}
1127 together with the identity \cite[p.449]{PBM}
1128 \begin{equation}
1129 \label{r5.26}
1130 F_{1}\left( a,b,b^{\prime };c,u,z\right) =(1-u)^{c-a-b}(1-z)^{-b^{\prime
1131 }}F_{1}\left( c-a,c-b-b^{\prime },b^{\prime },c,u,\frac{u-z}{1-z}\right)
1132 \end{equation}
1133 enable us to rewrite $I_{R}$ as
1134 \begin{equation}
1135 \label{r5.27}
1136 I_{R}=\frac{R}{(1-R\omega )^{2B}}F_{1}\left( 2-2B,1-2B,2B,2,R,\frac{%
1137 R-R\omega }{1-R\omega }\right)
1138 \end{equation}
1139 Therefore, the kernel function $\eqref{r5.16} $ reads
1140 \begin{equation}
1141 \label{r5.28}
1142 P_{R}^{B}(z,w)=\frac{(2B-1)^{2}R}{(1-R\bar{z}w)^{2B}}F_{1}\left(
1143 2-2B,1-2B,2B,2,R,\frac{R-R\bar{z}w}{1-R\bar{z}w}\right) .
1144 \end{equation}
1145 To check the limit of this kernel as $R\rightarrow 1$, we first observe that
1146 \begin{equation}
1147 \label{r5.29}
1148 F_{1}\left( 2-2B,1-2B,2B,2;R,\frac{R-R\bar{z}w}{1-R\bar{z}w}\right)
1149 \rightarrow F_{1}\left( 2-2B,1-2B,2B,2,1,1\right) \text{ as \ }R\rightarrow
1150 1.
1151 \end{equation}
1152 By using the identity \cite[p.452]{PBM}
1153 \begin{equation}
1154 \label{r5.30}
1155 F_{1}\left( a,b,b^{\prime },c;Z;Z\right) ={}_{2}F_{1}\left( a,b+b^{\prime
1156 },c;Z\right) ,
1157 \end{equation}
1158 the obtained expression limit in $\eqref{r5.29}$ reduces as
1159 \begin{equation}
1160 \label{r5.31}
1161 F_{1}\left( 2-2B,1-2B,2B,2B,2;1,1\right) ={}_{2}F_{1}\left(
1162 2-2B,1,2;1\right) .
1163 \end{equation}
1164 Finally, we use the Gauss theorem \cite[p.489]{PBM}:
1165 \begin{equation}
1166 \label{r5.32}
1167 {}_{2}F_{1}(a,b,c;1)=\frac{\Gamma (c)\Gamma (c-a-b)}{\Gamma (c-a)\Gamma (c-b)%
1168 }, \quad \rm{Re}(c-a-b)>0
1169 \end{equation}
1170 for parameters $a=2-2B,b=1$ and $c=2$ to get that
1171 \begin{equation}
1172 \label{r5.33}
1173 {}_{2}F_{1}\left( 2-2B,1,2;1\right) =\frac{1}{2B-1}.
1174 \end{equation}
1175 This, leads to the limit
1176 \begin{equation}
1177 \label{r5.34}
1178 \lim_{R\rightarrow 1}P_{R}^{B}(z,w)=\frac{(2B-1)}{(1-\bar{z}w)^{2B}}
1179 \end{equation}
1180 This completes the proof.
1181
1182 \smallskip
1183
1184 We end this section by observing that $\eqref{r5.8} $ may provides us
1185 with a family of Hilbert spaces $\left( \text{RKHS}\right) $ indexed by the
1186 continuous parameter $R\in \left] 0,1\right[ ,$ where each one would have $%
1187 P_{R}^{B}(z,w)$ as its reproducing kernel for which the Eq. $\eqref{r5.12} $ will represent a Zaremba expansion \cite{SFH}. These
1188 RKHS are natural generalizations of the Bergman space $\mathcal{A}^{B}(%
1189 \mathbb{D})$ and deserve to be investigated in details in a futur work.
1190
1191
1192 \begin{thebibliography}{99}
1193
1194 \bibitem{W} \bigskip M.W. Wong, Wavelet Transform and Localisation
1195 Operators, Operator Theory: Advances and Applications, Vol. 136, (2002) Birkh%
1196 \"{a}user, Basel
1197
1198 \bibitem{V} V.V. Dodonov, 'Nonclassical` states in quantum optics: a
1199 squeezed review of the first 75 years, J.Opt. B: Quantum Semiclass. Opt. 4
1200 (2002) R1-R33.
1201
1202 \bibitem{G} J.P. Gazeau, Coherent states in quantum physics, WILEY-VCH
1203 Verlag GMBH \& Co. KGaA Weinheim (2009)
1204
1205 \bibitem{TP} S. T. Ali, J. P. Antoine and J. P. Gazeau, Coherent States,
1206 Wavelets and Their Generalizations, second edition, Springer
1207 Science+Business Media New York (2014)
1208
1209 \bibitem{RS} M. Reed and B. Simon, Methods of Modern Mathematical Physics,
1210 Vols. I--IV \symbol{126}Academic, New York (1978)
1211
1212 \bibitem{Daub} I. Daubechies, Time-frequency localization operators: a
1213 geometric phase space approach, \textit{IEEE transactions on information
1214 theory}, Vol 34, N4 (1998)
1215
1216 \bibitem{PD1} D. Popov, Barut-Girardello coherent states for the
1217 pseudo-harmonic oscillator, J. Phys. A: Math. Gen. 34, 5283-5296 \ (2001)
1218
1219 \bibitem{PD2} D. Popov, Gazeau-Klauder quasi-coherent states for the Morse
1220 oscillator, Phys. Lett. A. vol 316 (6) "69-381 (2003)
1221
1222 \bibitem{Mou1} Z. Mouayn, Husimi's Q-function of the isotonic oscillator in a
1223 generalized Binomial states representation, \textit{Math. Phys. Anal. Geom}.
1224 17(3-4):289-303, (2014)
1225
1226 \bibitem{GJT} Ts Gantsog, A. Joshi and R. Tanas, Quantum Opt. vol 6 , 517-526
1227 (1994$)$
1228
1229 \bibitem{FS} H-C. Fu and R. Sasaki, Negative binomial states of quantized
1230 Radiation fields, arXiv:quant-ph/9610024v1
1231
1232 \bibitem{DMo} Duflo M. and Moore C.C., On the regular representation of a
1233 Nonunimodular locally compact group, \textit{J. Funct. Anal,} \textbf{21},
1234 2, pp.209-243 (1976)
1235
1236 \bibitem{AK} Aslaken E W and Klauder J R , J.Math.Phys 10, p.2267 (1969).
1237
1238 \bibitem{Mou2} Z. Mouayn, Characterization of hyperbolic Landau states by
1239 coherent state transform, J. Phys. A: Math \& Gen vol 36 (29) 8071 (2003$)$
1240
1241 \bibitem{MaOb} {W. Magnus, F. Oberhettinger, R.P. Soni}, \textit{\ Formulas
1242 and Theorems for the special Functions of Mathematical Physics}, 3rd edn.
1243 Die Grundlehren der mathematischen Wissenschaften, vol. 52. Springer, New
1244 York(1966)
1245
1246 \bibitem{BMMB} M. G. Benedict and B. Molnar, An algebraic construction of the
1247 coherent states of the Morse potential based on supersymmetric quantum
1248 mechanics,\textit{\ Phys. Rev}. 60 R 1737 (1999)
1249
1250 \bibitem{MP} P. M. Morse, Diatomic molecules according to the wave
1251 mechanics.II. Vibrational levels, Phys. Rev. D. 57 (1929)
1252
1253 \bibitem{NMSL} M. M. Nieto and L. M. Simmons Jr: Coherent states for general
1254 potentials. I Formalism, II. Confining one-dimensional examples; III Non
1255 confining one-diemensional examples, Phys. Rev. D. Vol 20, 1321,1332,1342 \
1256 (1979)
1257
1258 \bibitem{CDM} H. Chhaiba, N. Demni and Z. Mouayn, Analysis of generalized
1259 negative binomial distributions attached to hyperbolic Landau levels, \emph{%
1260 J. Math. Phys},\textbf{\ 57}, 072103 (2016)
1261
1262 \bibitem{IWGM} A. Intissar, F. El Wassouli, A. Ghanmi and Z. Mouayn.
1263 Generalized second bargmann transforms associated with the hyperbolic landau
1264 levels on the poincar disk. Annales Henri Poincar, 13 (4):513-524, 2012.
1265
1266 \bibitem{PBM} A. P. Prudnikov, Yu. A. Brychkov and O. I. Marichev. Integrals
1267 and Series-More special functions. Volume \textbf{3}. 1986.
1268
1269 \bibitem{SFH} F. H. Szafraniec, Reproducing kernel propoerty and its space :
1270 The basics , in D. Alpay (eds) Operator Theory, Springer, Basel (2015)
1271 \end{thebibliography}
1272
1273 \end{document}
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345