Advances in Pure Mathematics, 2012, 2, 169-176
http://dx.doi.org/10.4236/apm.2012.23023 Published Online May 2012 (http://www.SciRP.org/journal/apm)
Fock Spaces for the q-Dunkl Kernel
Fethi Soltani*
Higher College of Technology and Informati cs, Tunis, Tunisia
Email: fethisoltani10@yahoo.com
Received November 26, 2011; revised January 4, 2012; accepted January 20, 2012
ABSTRACT
In this work, we introduce a class of Hilbert spaces ,q
of entire function s on the disk 1
0,D


0< <1q
1q, , with
reproducing kernel given by the q-Dunkl kernel
2
;Ezq
,q
. The definition and properties of the space
,q
extend
naturally those of the well-known classical Fock space. Next, we study the multiplication operator Q by z and the
q-Dunkl operator
on the Fock space ,q
; and we prove that these operators are adjoint-operators and continu-
ous from this space into itself.
Keywords: Generalized q-Fock Spaces; q-D un kl Kernel; q-Dunkl Operator; q-Translation Operators
1. Introduction
Fock space (called also Segal-Bargmann space [1])
is the Hilbert space of entire functions

0
n
n
n
f
zaz
on such that
22
0
:!.
n
fan

n
This space was introduced by Bargmann in [2] and it
was the aim of many works [1]. Especially, the differen-
tial operator ddDz and the multiplication operator
by z are densely defined, closed and adjoint-operators on
(see [2]).
In [3], Sifi and Soltani introduced a Hilbert space
of entire functions on , where the inner product is
weighted by the modified Macdonald function. On
the Dunkl operator
   
d21
:,
d2
12,
f
zfz





fz fz
zz


and the multiplication by z are densely defined, closed
and adjoint-operators.
In this paper, we consider the q-Du nkl ker n el:

2
2
0,
;
n
n
x
bq
;:
n
Ex
q

2
;
n
bq
where are given later in Section 2. We dis-
cuss some properties of a class of Fock spaces associated
to the q-Dunkl kernel and we give some appl ications.
In this work, building on the ideas of Bargmann and
Cholewinski [4], we define the q-Fock space ,q

0
n
n
n
as
the space of entire functions zaz
on the
f
disk 1
0,1
Dq



of center 0 and radius 1
1q
, and
such that

,
22 2
0
:;
qnn
n
fabq
.

Let f and g be in ,q

n
, such that 0n
n
f
zaz

0
n
n
n
and zcz
, the inner product is given by

,
2
0
,;.
qnnn
n
gacbq
,q
f
The q-Fock space
,q
has also a reproducing kernel

given by

2
,1
,;;,0,.
1
qwzE wzqwz Dq





,q
f
Then if

, we have

,
,1
,,., 0,.
1
q
q
fw fwwDq




,q
Using this property, w e prove that the space
,q
is a
Hilbert space and we give an Hilbert basis.
Next, using the previous results, we consider the mul-
tiplication operator Q by z and the q-Dunkl operator
*Author partially supported by DGRST project 04/UR/15-02 and CMCU
on the Fock space ,q
, and we prove that these
p
rogram 1 0 G 1 50 3 .
C
opyright © 2012 SciRes. APM
F. SOLTANI
170
operators are continuous from ,q
into itself, and sat-
isfy:
,
,,
,
q
qq
fCf

,q

,,
,,
qq
q
C f


C
Qf
where ,q
is a constant independent of f.
Then, we prove that these operators are adjoint-op-
erators on ,q
:
,,
,,
qq
Qf gfg
,,
;,.
qq
f g
,q

Lastly, we define and study on the Fock space
,
the q-translation operators:



2
,
:;
;
zq
Tf wEzqf wwz

 1
,0,,
1
Dq



and the generalized multiplication operators:



2
:; ;
z
MfwE zQqfwwz

1
,0,.
1
Dq



q
Using the continuous properties of ,
and Q we
deduce also that the operators
z
T and
z
M
, for


21
1
12qq
 ,q
z, are continuous from
into
itself, and satisfy:

,,
2
;,
qq
z qf


,zq
TfE C


,,
2
;.
qq
zq f


,
q
,zq
MfE C

2. Preliminaries and the q-Fock Spaces
Let a and q be real numbers such that ; the
q-shifted factorial are defined by 0< <1q
11,2,,.
n
n


 

00
;:1, ;:1,
i
n
i
a qaqaq

Jackson [5] defined the q-analogue of the Gamma
function as



1
;
:1
;
x
qx
qq
xq
qq
 ,0,1,2,x
It satisfies the functional equation
 
1
1
qq
q
xx
q
 

1,
11,
x
q
 

and tends to
x
2,

when q tends to 1. In particular, for
, we have
1,n

;
1.
1
n
n
qq
q
q
Df

qn
The q-derivative of a suitable function f (see
[6]) is given by


:,0,
1
q
fx fqx
Df xx
qx
00Df f

0f

q
Df x
and q provided exists.
If f is differentiable then tends to
f
x
as
1q
 


.
Taking account of the pap er [3] and the same way, we
define the q-Dunkl kernel by

2
22 2
1
;: ;;,
11
q
x
ExqIxqI xq
q
 


2
;
I
where xq
is the q-modif ied Bessel function [7,8]
given by

 
2
22
2
2
2
0
;
:1 .
111
n
n
qnqq
Ixq
x
qn n
 

2
;Exq
 
Furthermore, the Dunkl kernel can be ex-
panded in a power series in the form
2
2
0
;: ,
;
n
nn
x
Exq bq

 

(1)
where
22
2
2
2
2
111
;: ,
1
n
qq
n
q
qn n
bq



 

and
22
2
21
2
21
112
;: .
1
n
qq
n
q
qnn
bq



If we put 2
1
:;
n
n
Ubq
, then

2
1
1,1.
1
n
n
Uq
Uq


2
;Exq

Thus, the q-Dunkl kernel is defined on
2
1
0, 1
D
q




and tends to the Dunkl kernel Ex
1q
as
,q
.
We consider the q-Dunkl operator operator
de-
fined by
  
 
,
21
:,
2
q
qq
f
qxf qx
fx Dfxx



where
Copyright © 2012 SciRes. APM
F. SOLTANI 171

21
1
: .
1
q
q

21
q
The q-Dunkl op erator tends to the Dunkl oper ator
as .
1q
Lemma 1. The function
2
.;Eq
, 1
0,1
Dq
,
is the unique analytic solution of the q-problem:
,qyx yx
,01.y
0
n
n
nax

1
1.
n
qn
q
a nx
(2)
Proof. Searching a solution of (2) in the form
. Then

yx

n
Dy x
Replacing in (2), we obtain
 

1
11
.
n
n
nn
nn
ax




11
21 2
n
n
qq
an qx










Thus,
 

1
11
21 2
n
n
nn
qq
an qa






 




, 1,2,.n
Using the fact that
21
21 21
n
qq
nq
22 2
q
n
 , we deduce that
 
22121
and
22
nnn
aaa
nn


2
.
22
n
qq
a


We get

2
21 22nn
21
.
22
nn
qq
aa


Since

2
21
qq
nqn
, we deduce

 
22
21
.
11
nn
qq
aa


 
2
21 2
qn n

This proves that
 
2
22
21
21 21
11
n
q
nn
qq
a
qn



1,
2
n

and
 
2
2
22
22
1
111
n
q
nn
qq
a
qn n



 
 
.
Therefore,
 
 




2
22
2
22
2
2
2
0
21
21
0
22
1
1111
1112
;;,
11
n
n
qnqq
n
n
qnqq
q
x
yx
qn n
x
qnn
x
IxqI xq
q


 

 



2
;
n
bq
n
which completes the proof of the lemma.
Lemma 2. The constants , satisfy
the following relations:
2
1;
n
bq


1)

12
11
121 ;,
2
n
n
n
qq
nq bq






 





2
22
21 2
;1 11;
nn
q
bq qb q

 , 2)
2
;
n
bq

3)


22
2
1
121 1
2
:(1)
n
qq
q
n
qn


 




 ,
where 2n is the integer part of 2n
k
.
Lemma 3. For , we have

2
1
,2
1
;,1.
;
k
kk
q
k
bq
zzk
bq




Proof. Since
2
2
0
;: ,
;
k
kk
z
Ezq bq

then from Equat ion (2 ) we ob tai n

1
,
22
11
1
.
;;
kk
qkk
kk
kk
zz
bqb q




This clearly yields the result.
Definition 1. Let 12
 ,q
. The q-Fock space

0
n
n
n
is the prehilbertian space of entire functions
zaz
f
on 1
0,1
Dq



, such that

,
2
22
0
:;,
qnn
n
fabq

(3)
2
;bq
,q
where n
The inner product in
is given by (1).
is given for

0
n
n
n
zaz

0
n
n
n
zcz
by
f
g
and
Copyright © 2012 SciRes. APM
F. SOLTANI
172

2
,;.
nnn
,0
qn
f
ga
cbq
1q
(4)
Remark 1. If , the space ,q
,
,q
fg
agrees with
the generalized Fock space associated to the Dunkl
operator (see [ 3]).
Proposition 1. For
, we have

 
,,
,0
qq,.
f
gfg

gzgz

n
az
Proof. Given ,
0nq
n

fz

,
0
n
nq
n
gz cz and


. Since from Lemma 3,


2
,2
;bq
,,
;
k
nk kn
q
kn
zz
kn
bq
 


(5)
we can write

,
2
0
;
n
qn
nn
g
0.
g
zz
bq
(6)
Using (4) and (6), we get
 
,,
0.
nn
nq
a g

,00
,0
qnq
nn
fga g


 

Thus

,
,0,
qg

,q
,q
fgf
which gives the desired result.
The following theorem proves that
is a repro-
ducing kernel space.
Theorem 1. The function ,q
given for
1

,0,
1
Dq


wz , by


2
,;,E wzq

,q
,
qwz
is a reproducing kernel for the q-Fock space
, that
is:
1) For all 1
0,1
wD q



,q
,q
,zw
, the function
z
belongs to
.
2) For all 1
wD


 ,q
f
0,1q
and
, we have
 
,
. .
q
,
,,
q
f
wfw
Proof. 1) Since


,0
1
,0
,
1
q
n
wz
zw Dq
2
,;
;
,
n
n
n
wz
bq



then from (3), we deduce that

(7)


,
22
,2
0
,.; ,
;
q
q
nn
wE
wq
bq

2
2n
w

which p roves 1) .
,
0
n
nq
naz
2) If

fz
, from (4) and (7), we
deduce

,
,0
1
., 0,.
1
q
n
qn
n
awfwwDq




This completes the proof of the theorem.
,,fw
Remark 2. From Theorem 1 2), for ,q
f
and
1
0,D


, we have
1
wq


,
,
,
,
12
22
,.
;.
q
q
q
q
fww f
Ewq f



(8)
Proposition 2. The space ,q
equipped with the in-
ner product ,
.,.q
is an Hilbert space; and the set
2
.;q
nn
given
 
by
2
2
1
;,0,,
1
;
n
n
zq z Dq
bq




forms an Hilbert basis for the space ,q
n
z
.
Proof. Let
nn
f
be a Cauchy sequence in ,q
.
,
lim, in.
nq
n
ff
We put

From (8), we have
 
,
12
22
;.
q
npnnpn
Ewqf f



This inequality shows that the sequence
fwfw


nn
f is
pointwise convergent to f. Since the function
12
22
;wEwq
is continuous on 1
0,1q



,
then
D
nn
f
convergall compac
es to f uniformly on t
set of1
. Consequently, 0,1
Dq

f
is an entire funct io n
on 1
0,1
Dq



, then
f
belongs to the space ,q
.
n the other omt Ohand, fr the relation (4), we ge
,q
where ,nm
22 ,
.; , .;,
nm nm
qq
 
is the Kronecker symbol.
sho
n is an or-
thonormal set in ,q
This ws that the family
2
.;
nq
.
Copyright © 2012 SciRes. APM
F. SOLTANI 173
Let

0
n
n
n
f
zaz
be an,q
element of
such
that

,0, .
q
n
2
,.;
n
fq

From the relation (4), we deduce that
0, .an
Remark 3.
n
T
his completes the proof.
1) The set

2
.; ,1
0,Ewq wD





nse in
1q

is de,q
.
2) For all 1
,0,
1
Dq


, we have zw
 
,
222
;.; .
q
zq Ezq
3. Operators on the Fock Spaces ,
q
,.;Ew Ewq

On ,q
, we consider the mcation operators Q and
  
ultipli
q
N given by
 
:,Qfzzf z
 
:.
qq 1
f
zfqz
Nf zzDf zq

We den also by ,q
ote
the q-Dunkl o peratoefined
for enti
r d
re functions on1
0,D

.
,,,
1q

We write
,.
q
Q
qq
Q Q


culation we obtain.
,,qqq
QBW

ightforward calThen by stra
Lemma 4. ,


 , wher e

:,
q
Bfzf qz
 
q
,:1
1.
2
q
qqq
WqBB




Remark 4. The Lemma 4 is the analogous commuta-
tion rule of [3]. When 1q
21
, then ,,
qQ


tends to

21
I
B, whe
 re tity operator and B is
hI is the iden
te parity operator given by

:Bf zfz
.
Lemma 5. If ,q
f
q
Bf, qd ,q
Wf
,q
thenNf an
belong to
, and
1) ,q
q
,,qq
q
BffB f


,
2) ,
1q
f
q
,q
q
Nf 1
3)
,
,
21 q
qf
,
,q
q
Wf
Proof. Let

fz
 .
,
0
n
nq
az
n

, then
 
0,
nn
qn
n
Bfzfqzqz


a


0,
1
n
qn
n
fz fqz
Nf zanz
q

and from (3), we obtain
q

,0
qn

,
2222
22
2
0
;
;,
q
n
qnn
nn
n
Bfaqbq
ab qf

and



,
2
222
0;.
q
qnn
q
n
Nfan bq
Using the fact that

1
1
q
nq
, we deduce

 
,
222
0
11
.
q
n
a bf
On the other hand from 1) we deduce that
,
2
22
;
11
q
qn
n
Nf q
qq



 

,
,,
,
,
11
2
21 ,
q
qq
q
q
qq
q
Wf
qBfqBf
f

21
q



which completes the proof of the Lemma.
We now study the continuous proper ty of the oper ators
,q
and Q on ,q
.
Theorem 2. If ,q
f
then ,q
f
and Qf belong
to ,q
, and we have
1) ,
,q
q
,,qq
fCf
2)

 ,
,
,
q
q
QfCf ,q

, where
12
,1
:21 .
q
C




Proof. Let

,
0
n
nq
n
fz az
1
qq

.
1) From Lemma 3,





2
1
,11
2
1
12
0
;
;
;.
;
nn
qn
nn
nn
n
nn
bq
f
2
za z
bq
bq
az
bq
(9)
Then from (9), we get


2
21
22
,1 1
2
;;.
;
n
qn n
n
bq
f
,0
qn
abq
bq


Using Lemma 2 1), we obtain
Copyright © 2012 SciRes. APM
F. SOLTANI
174
 


,
2
,
2
0;.
2
q
q
n
n
n
f
bq




(10)
Using t

211
21
n
n
qq
an
q


 

he fact that 1
1
q
nq
, we obtain

,
,
2
,,
0
,.
q
q
qq
n
q
fC
Cf


12
2
;
nn
abq



2) On the other hand, since
1
1
,
n
n
n
a z
(11)
th

Qf z
en
 
0;.
nn
a bq
,
22
222
11
=1 =
=;=
qnn
nn
Qfa bq




emma 2 1), we deduce By L
 


,
2
2
11 ;.
2
q
n
n
q
Qf
bq







(12)
Using t

21
012
1
n
n
q
n
an q

he fact that 1
11
q
nq

, we obtain
,,
,.
qq
q
C f


following norm
If ,q
f
Qf
We deduce also the equalities.
Theorem 3.
then
1) ,,qq
2
,,
qq
ffNf



,
,,
2q
qq
B f

21
qfB
2) ,,
2,
qq
q
Qff N f
,
2
q
q
B f


,q
qq
B f
21 ,
2
q
q
fB

,
3) ,,
2
qq
q
QffB f
,,
2
2
,,
,
qq
qq
f Wf


.
,
0nq
az
 
Proof. Let

n
n
fz
.
rom (10), we get 1) F




,
,,
2
,
2
11
21 ;
2
21
,,.
2
q
qq
q
n
n
n
qq
q
qqq
f
nq bq
fNffBBf








 
2) On the other hand, by (12) and using the fact t
2
n
a
0n
hat
1n
qq
nnq, we ob tain

 






,
,,
,
22
222
00
2
;;
21
,
21 ,.
2
q
qq
q
n
nn nn
q
nn
qq
q
qq
Qfanbq aqbq
q
fNfB f
q
fB Bf







3) Follows directly from 1) and 2).
Proposition 3. The operators Q and ,q
22
011;
2
n
qn
nn
n
aq b q

are ad-
joint-operators on ,q
; and for all ,
,q
fg
, we
have
,,
qq
,
,,.
q
Qf gfg

Proof. Consider

0
n
n
n
f
zaz
and

0
n
n
n
g
zcz
in,q
),



. From (9) and (11
2
1
,1
2
0
;,
;
n
qn
nn
bq
n
g
zc z
bq

an

1
1
.
n
n
n
Qf zaz
Thus from (4), we get
d


,
,
2
1
1
2
11 ,
,;
;, ,
q
q
nnn
n
nn nq
gacbq
ac bqfg


which gives the result.
In the next part of this section we study a generalized
translation and multiplication operators on ,q
0n
Qf
. We
begin by the following defi nition.
Definition 2. For ,q
f
and 1
,0,wz D


1q
we define:
,
The q-translation operators on ,q
, by
Copyright © 2012 SciRes. APM
F. SOLTANI 175


 

,
2
,2
0
:; ;
n
qn
zq
n
fw
Tf wEzqf wbq

 
(13)
.
n
z
The generalized m ultiplication operators on ,q
, by





2
02.
;
n
n
Qfw
:;
z
nn
M
fw
E zQqfwz
bq

(14)
For 1
,0,
1
Dq




2
.;Eq
product
wz
tisfies the following
, the function
formulas:
sa-



222
.;;; ,TEqwEzqE wq
z



 
22
; .
2
.; ;
z
M
EqwEwz

Remark 5. If 1q
in (13),
qEwq
we obtain the general-
slation operators 102).
,
n
nq
az
ized trangiven in ([9], page
Proposition 4. Let 0n
fz

 and
1
,
1
,0
Dq


. Then zw 
1)



2
;
n
nnkk
n
n
bq
Tf waz
 ,
00 22
;;
zk
kn
k
bq
b q

with
w




22
22
2
;
n
bq
q
22
2
;;
1
,; 1 1
2,
11
11
22
kn
k
qq
qq
bq
b
n
cnkq
knk










  
 


 
 

where



2
22
2
,;
21
q
qq
n
cnkq
k


21 .
1
2
nk

 





2)

00
n
znk
k
a
2
;
kn
nk
M
fw


Proof. Let

n
z w
bq




.
,
0nq
az
n
fz
.

1) From (13), we have


,
2
0;,
;
1
0,.
1
qn
z
nn
fw
Tf wzwzDq
bq




from (5) , we ha ve


n
But

2
;.
;
k
nkn
kn kn
bq
fw aw
bq




,2
qk
Thus we obtain
2
22
00
;.
;;
nnnkk
zn
nk
knk
bq
Tf wawz
bqb q



Oemma 2 3), we get
n the other hand from L



22
2
22
2
;
1
,; 11
2,
11
1
n
qq
bq
q
n
cnkq
knk








  
 
which gives the 1).
2) From (14), we have

22
1
22
qq




 
 

;;
kn
k
bq
b


2
0
1
;, 0,
1
;
n
n
z
nn
Qfw
Mfwzwz Dq
bq




But from (11), we have

.
kn
kn
f waw


.
nk
Q
Thus we obtain
2
00 .
;
nkn
nk
z
nk k
a
M
fwz w
bq





According to Theorem 2 we study the continuous
property of the operators

z
T and
M
on ,q
z
.
,q
Theorem 4. If f
and


21
1
12
zqq
 ,
then z
Tf and zf belong to ,q
M
, and we ha
1)
ve
,,
2
,;
qq
zq
TfECz qf


,
2)
,,
2
,;
qq
zq
MfE Czqf

,
where ,q
C
are thnts of Te constaheorem 2.
Proof. From (13) and Theorem 2 1), we deduce



,,
,
,2
0
,
2
;
qq
q
zq
nn
n
q
Tf fbq
Cz f
bq

0nn
||
n
nz
;
.
Since


21
12qq


1
z, then ,2
1
1
qq
,
and therefore
||Cz
,,
2
,;,
qq
zq
TfECz qf



which gives the first inequality, and as in the same way
Copyright © 2012 SciRes. APM
F. SOLTANI
Copyright © 2012 SciRes. APM
176
we prove the second inequality of this tm.
From Proposition 3 we deduce the following results.
Proposit ,
,q
fg
heore
ion 5. For all
, we hav e
,,
,,,
qq
zz
fTg


Mfg
,,
,,.
qq
zz
f Mg


note by
Tf g
We de
z
R the folloerator defined onwing op
,q
by
 
 
22
,
;;
.
q
Ez
QqEzq


Then, we prove the following theorem.
22
,
:;;
zzz zzq
RTMMTEz qEzq


,q
Q
Theorem 5. For all f
, we have
,,
22
qq ,
,.
zz z
Mf TffRf
 


Proof. From Proposition 5, we get
q

,, ,
2
qq q
zzz zzz


,,
2,,
,.
qq
zz
MffTMff MTRf
TffRf




REFERENCES
[1] ators on
of the Ameri-
ciety, Vol. 1987, pp. 813-
829. doi:10.1090/S0002-9947-1987-0882716-4
C. A. Berger and L. A. Coburn, “Toeplitz Oper
the Segal-Bargmann Space,” Transactions
can Mathematical So 301, No. 2,
[2] V. Bargmann,ytic Functions “On a Hilbert Space of Anal
and an Associated Integral Transform, Part I,” Commu-
nications on Pure and Applied Mathematics, Vol. 14, No.
3, 1961, pp. 187-214. doi:10.1002/cpa.3160140303
[3] M. Sifi and F. Soltani, “Generalized Fock Spaces and
Weyl Relations for the Dunkl Kernel on the Real Line,”
Journal of Mathematical Analysis and Applications, Vol.
270, No. 1, 2002, pp. 92-106.
doi:10.1016/S0022-247X(02)00052-5
F. M. Cholewinski, “Generalized Fock Spaces and [4] Asso-
ciated Operators,” SIAM Journal of Mathematical Analy-
sis, Vol. 15, No. 1, 1984, pp. 177-202.
doi:10.1137/0515015
[5] G. H. Jackson, “On a q-Definite Integra
Journal of Pure and Applied Mathema
ls,” The Quarterly
ns, Vol. 14, 1997,
tics, Vol. 41, No. 2,
1910, pp. 193-203.
[6] T. H. Koornwinder, “Special Functions and q-Commuting
Variables,” Fields Institute Communicatio
pp. 131-166.
[7] A. Fitouhi, M. M. Hamza and F. Bouzeffour, “The q-jα
Bessel Function,” Journal of Approximation Theory, Vol.
115, No. 1, 2002, pp. 144-166.
doi:10.1006/jath.2001.3645
[8] F. Soltani, “Multiplication and Translation Operators on
the Fock Spaces for the q-Modified Bessel Function,” The
Advances in Pure Mathematics (APM), Vol. 1, No. 2,
2011, pp. 221-227. doi:10.4236/apm.2011.14039
[9] J. J. Betancor, M. Sifi and K. Trimèche, “Hypercyclic and
Chaotic Convolution Operators Associated with the Dunkl
Operator on ,” Acta Mathematica Hungarica, Vol. 106,
No. 1-2, 2005, pp. 101-116.
doi:10.1007/s10474-005-0009-1