• Aucun résultat trouvé

On estimates for the Fourier transform in the space Lp(Rn)

N/A
N/A
Protected

Academic year: 2021

Partager "On estimates for the Fourier transform in the space Lp(Rn)"

Copied!
6
0
0

Texte intégral

(1)

DOI 10.1007/s13370-014-0278-3

On estimates for the Fourier transform in the space L p ( R n )

R. Daher · M. El Hamma

Received: 9 November 2013 / Accepted: 20 June 2014

© African Mathematical Union and Springer-Verlag Berlin Heidelberg 2014

Abstract In this paper, we prove two estimates useful in applications for the Fourier trans- form in the space L

p

(

Rn

), 1 < p ≤ 2, as applied to some classes of functions characterized by a generalized modulus of continuity.

Keywords Fourier transform · Generalized modulus of continuity Mathematics Subject Classification 42B12

1 Introduction and preliminaries

The Fourier transform, as well as Fourier series, is widely used in various fields of calculus, mathematical physics, etc.

In [1], Abilov et al. proved two estimates for the Fourier transform in the space of square integrable functions on certain classes of functions characterized by the generalized mod- ulus of continuity. In this paper, we prove these estimates in the space L

p

(

Rn

) of p-power integrable functions. We prove out that similar results have been established in the context of Bessel transform and Jacobi transform [2,3].

Assume that L

p

(

Rn

), (1 < p ≤ 2), is the space of p-power integrable functions f :

R

−→

C

with the norm

f

p

=

Rn

| f (y)|

p

d y

1/p

.

R. Daher·M. El Hamma (

B

)

Department of Mathematics, Faculty of Sciences Aïn Chock, University of Hassan II, Casablanca, Morocco

e-mail: [email protected] R. Daher

e-mail: [email protected]

(2)

The Fourier transform for the function f ∈ L

p

(

Rn

) , 1 < p ≤ 2, is defined by f (ξ) = 1

( 2 π)

n/2

Rn

f ( x ) e

−iξ.x

d x The inversion formula of Fourier transform is defined by

f ( x ) = 1 ( 2 π)

n/2

Rn

f (ξ) e

iξ.x

d ξ

The Fourier transform above extends to a bounded linear map f −→ f from L

p

(

Rn

) to L

q

(

Rn

) , for 1 < p ≤ 2 and

1p

+

1q

= 1 (see [5]), so

f

q

C f

p

, (1)

where C is a positive constant and f ∈ L

p

(

Rn

).

In this paper, we estimate the integral

|λ|≥N

| f (λ)|

q

d λ,N > 0 . in certain classes of functions in the space L

p

(

Rn

).

Below, to simplify the calculations, we consider only two variable functions. Similar results are also valid for multivariable functions.

In L

p

(

R2

) , we consider the operator F

h

f (x, y) = 1

4h

2

x+h

x−h

y+h

y−h

f (ξ, η)dξd η, h > 0.

which is analogous to Steklov’s function.

We define the differences of first and higher orders as follows

h

f ( x , y ) = F

h

f ( x , y )f ( x , y ) = ( F

h

− E ) f ( x , y )

kh

f ( x , y ) =

h

(

k−1h

f ( x , y )) = ( F

h

− E )

k

f ( x , y ) =

k i=0

(− 1 )

k−i

k i

F

ih

f ( x , y ), (2) where

F

0h

f (x, y) = f (x , y), F

ih

f (x , y) = F

h

(F

i−1h

f (x , y)), i = 1, 2, . . . , k; k = 1, 2, . . . and E is the unit operator in the space L

p

(

R2

).

The quantity

k

( f, δ) = sup

0<h≤δ

kh

f (x, y)

p

is called the generalized modulus of continuity of kth order of the function f ∈ L

p

(

R2

) . Let W

r,kp

(D) denote the class of functions f ∈ L

p

(

R2

) having the generalized partial derivatives

f

∂x ,

2

f

∂x∂y , · · ·

in the sense of Levi (see [4]) that belong to f ∈ L

p

(

R2

), such that

k

(D

r

f, δ) = O(φ(δ

k

)),

(3)

where

D =

2

∂x

2

+

2

∂y

2

,

D

0

f = f , D

i

f = D(D

i−1

f ), i = 1, 2, . . . , r , and φ(t ) is an arbitrary function defined on [0, ∞).

Since

F

h

f (x , y) = 1 2π

−∞

−∞

sin(hξ) sin(hη)

hξhη f (ξ, η)e

i(xξ+ηy)

dη.

Lemma 1.1 Let f ∈ W

r,kp

( D ) . Then

−∞

−∞

2

+ η

2

)

qr

1 − sin ( h ξ) sin ( h η) h ξ h η

qk

| f (ξ, η)|

q

d ξ d ηC

q

kh

D

r

f ( x , y )

qp

.

Proof We have

(D

r

f )(ξ, η) = (−1)

r

2

+ η

2

)

r

f (ξ, η).

Then

(F

ih

D

r

f )(ξ, η) = (−1)

r

2

+ η

2

)

r

sin(hξ) sin(hη)

i

f (ξ, η).

From formula (2), we conclude that the Fourier transform of

kh

D

r

f ( x , y ) is (− 1 )

r

2

+ η

2

)

r

1 −

sin(hξ)hξhsin(hη)η

k

f (ξ, η). By formula (1), we have the result.

2 Main result

In this section we give the main result of this paper.

Theorem 2.1 For functions f (x , y) ∈ L

p

(

R2

), 1 < p2, in the class W

rp,k

(D),

sup

Wr,kp,φ(D) ξ22N2

| f (ξ, η)|

q

d ξ d η

1/q

= O

N

−2r

φ( c N )

k

,

where

1p

+

1q

= 1, c > 0 is a fixed constant and φ(t) is an arbitrary function defined on the interval [ 0 , ∞) .

Proof Let f ∈ W

r,kp,φ

(D). Taking into account the Hölder inequality yields

(4)

ξ22≥N2

| f (ξ, η)|

q

dξd η

ξ22≥N2

sin(hξ) sin(hη)

| f (ξ, η)|

q

d ξdη

=

ξ22N2

1 − sin ( h ξ) sin ( h η) h ξ h η

| f (ξ, η)|

q

d ξ d η

=

ξ22≥N2

1 − sin(hξ) sin(hη)

| f (ξ, η)|

q1/k

| f (ξ, η)|

1/k

ξ22≥N2

| f (ξ, η)|

q

qk−1

qk

ξ22≥N2

1 − sin(hξ) sin(hη) hξhη

qk

| f (ξ, η)|

q

1

qk

=

ξ22≥N2

| f (ξ, η)|

q

qk−1

qk

×

ξ22≥N2

1

2

+ η

2

)

qr

2

+ η

2

)

qr

1 − sin(hξ) sin(hη)

qk

| f (ξ, η)|

q

d ξdη

1

qk

N

−2r/k

ξ22≥N2

| f (ξ, η)|

q

qk−1

qk

×

ξ22N2

2

+ η

2

)

qr

1 − sin ( h ξ) sin ( h η) h ξ h η

qk

| f (ξ, η)|

q

d ξ d η

1

qk

From Lemma 1.1, we have the inequality

ξ22≥N2

2

+ η

2

)

qr

1 − sin(hξ) sin(hη) hξhη

qk

| f (ξ, η)|

q

d ξ d ηC

q

kh

D

r

f ( x , y )

qp

. Therefore

ξ22≥N2

| f (ξ, η)|

q

d ξdη

ξ22≥N2

sin(hξ) sin(hη)

| f (ξ, η)|

q

+ C

1/k

N

2r/k

ξ22N2

| f (ξ, η)|

q

d ξ d η

qk−qk1

kh

D

r

f ( x , y )

1/kp

Let

I =

ξ22≥N2

sin(hξ) sin(hη)

hξhη | f (ξ, η)|

q

From [1], we have

|I| ≤

√ 2 N h

ξ22≥N2

| f (ξ, η)|

q

dξdη Consequently

ξ22≥N2

sin(hξ) sin(hη)

hξhη | f (ξ, η)|

q

≤ 4 √

2 N h

ξ22≥N2

| f (ξ, η)|

q

dξd η.

(5)

Then

ξ22N2

| f (ξ, η)|

q

d ξ d η ≤ 4 √ 2 N h

ξ22N2

| f (ξ, η)|

q

d ξ d η + C

1/k

N

−2r/k

ξ22≥N2

| f (ξ, η)|

q

d ξ d η

qk−1

qk

kh

D

r

f ( x , y )

1p/k

Setting h =

Nc

in the last inequality and choosing c > 0 such that 1 −

4c2

12

, we obtain 1 − 4 √

2 c

mat hop

ξ22N2

| f (ξ, η)|

q

d ξ d η

C

1/k

N

−2r/k

ξ22N2

| f (ξ, η)|

q

d ξ d η

qk−1

qk

kh

D

r

f ( x , y )

1/p k

Hence

ξ22≥N2

| f (ξ, η)|

q

dξd η = O( N

2r q

kc/N

D

r

f (x, y)

qp

) Since

kc/N

D

r

f (x , y)

p

= O

φ c N

k

, we obtain

ξ22≥N2

| f (ξ, η)|

q

d ξ d η

1/q

= O

N

−2r

φ c N

k

which proves Theorem 2.1.

Corollary 2.2 Let f ∈ W

rp,,ktν

(D), (ν > 0), then

ξ22≥N2

| f (ξ, η)|

q

= O(N

−2qr−qkν

) where r = 0 , 1 , 2 , . . . ; k = 1 , 2 , . . .

Proof Let f ∈ W

rp,,ktν

(D) and φ(t) = t

ν

. Then From Theorem 2.1, we have

ξ22≥N2

| f (ξ, η)|

q

d ξ d η = O ( N

−2qr−qkν

) Thus, the proof is finished.

References

1. Abilov, V.A., Abilova, F.V., Kerimov, M.K.: Some Remarks Concerning the Fourier Transform in the Space L2(Rn). Comput. Math. Math. Physics 48(12), 2146–2153 (2008)

2. R. Daher and M. El Hamma, On Estimates for the Bessel Transform in the Space Lp,α(R+)Thai Journal of Mathematics Vol (11) (2013) No. 3. pp. 697–702.

3. Daher, R., El Hamma, M.: On Estimates for the Jacobi transform in the Space L2(R+, (α,β)(t)dt)Inter.

J. of Applied Mathematics. 25(1), 13–23 (2012)

(6)

4. Nikol’skii, S.M.: Approximation of Functions of Several Variables and Embedding Theorems. Nauka, Moscow (1969). [in Russian]

5. Stein, E.M., Weiss, G.: Introduction to Fourier analysis on Euclidean spaces, Princeton Univ. Press, Prince- ton N. J (1971)

Références

Documents relatifs

When the symbol is continuous on the closed unit disk and has a domain touching the boundary non-tangentially at a finite number of points, with a good behavior at the boundary

In particular, estimates (3.10), (3.11) and (3.12) (with ν = 1) should be compared with the results on the monochromatic inverse scattering problem obtained by H¨ ahner, Hohage

The quadratic phase applied in the spectral domain can be actually interpreted as a delay applied to the different spectral components of the modulated signal: the low

Proposition 4.6. The remains of this proof is identical to the one of Corollary 2.2. We introduce the following lemma... Similarly to Corollary 2.4, we have the following result...

The generalized Dunkl translations are constructed on the base of certain differential-difference operators which are widely used in mathematical physics (e.g., [5, 9])..

The theory of Dunkl operators provides generalizations of var- ious multivariable analytic structures, among others we cite the exponential function, the Fourier transform and

Abstract In this paper, using a generalized translation operator, we obtain an analog of Younis’s Theorem 5.2 in Younis (Int J Math Math Sci 9:301–312, 1986) for the

Titchmarsh’s [5, Therem 85] characterized the set of functions in L 2 ( R ) satisfying the Cauchy Lipschitz condition by means of an asymptotic estimate growth of the norm of