The aim of this work is to study the existence of a periodic solution for some neutral partial functional differential equations. Our approach is based on the R-boundedness of linear operators L
p-multipliers and UMD-spaces.
Motivated by the fact that neutral functional differential equations (abbreviated, NFDE) with finite delay arise in many areas of applied mathematics, this type of equations has received much attention in recent years. In particular, the problem of existence of periodic solutions has been considered by several authors. We refer the readers to papers [1] - [8] and the references listed therein for information on this subject.
In this work, we study the existence of periodic solutions for the following neutral partial functional differential equations of the following form
where is a linear closed operator on Banach space and for all. For (some) L and G are in is the space of all bounded linear operators and is an element of which is defined as follows
In [4] , Ezzinbi et al. established the existence of periodic solutions for the following partial functional differential equation:
where is a continuous w-periodic function, is a con- tinuous function w-in t, periodic and G is a positive function.
In [1] , Arendt gave necessary and sufficient conditions for the existence of periodic solutions of the following evolution equation.
where A is a closed linear operator on an UMD-space Y.
In [2] , C. Lizama established results on the existence of periodic solutions of Equation (1) when namely, for the following partial functional differential equation
where is a linear operator on an UMD-space X.
In [3] , Hernan et al., studied the existence of periodic solution for the class of linear abstract neutral functional differential equation described in the following form:
where and are closed linear operator such that
and.
The organisation of this work is as follows: In Section 2, we present preliminary results on UMD spaces. In Section 3, we study the existence of periodic strong solution for Equation (1) with finite delay and we discuss the existence of mild solutions of Equation (1). In Section 4, we give the main abstract result [Theorem 4.1] of this work, and some important consequence when A generates a -semigroup [Theorem 4.2]. The last section is devoted to some examples.
2. UMD Spaces
Let X be a Banach space. Firstly, we denote By the group defined as the quotient. There is an identification between functions on and 2p-periodic func- tions on. We consider the interval as a model for.
Given, we denote by the space of 2p-periodic locally p-inte- grable functions from into X, with the norm:
For, we denote by, the k-th Fourier coefficient of f that is defined by:
Definition 2.1 Let and. Define the operator by: for all
if exists in Then, is called the Hilbert transform of f on.
Definition 2.2 [2]
A Banach space X is said to be UMD space if the Hilbert transform is bounded on for all.
Example 2.1 [9] 1) Any Hilbert space is an UMD space.
2) (0.1) are UMD spaces for every.
3) Any closed subspace of UMD space is an UMD space.
R-Bounded and Lp-Multipliers
Let X and Y be Banach spaces. Then denotes the space of bounded linear ope- rators from X to Y.
Definition 2.3 [1]
A family of operators is called R-bounded (Rademacher bounded or randomized bounded), if there is a constant and such that for each, and for all independent, symmetric, -va- lued random variables on a probability space the inequality
is valid. The smallest C is called R-bounded of and it is denoted by.
Lemma 2.1 ( [2] , Remark 2.2)
1) If is R-bounded then it is uniformly bounded, with
2) The definition of R-boundedness is independent of
Definition 2.4 [1] For, a sequence is said to be an -multiplier if for each, there exists such that
for all.
Proposition 2.1 ( [1] , Proposition 1.11) Let X be a Banach space and be an -multiplier, where. Then the set is R-bounded.
Theorem 2.3 (Neumann Expansion) Let, where X is a Banach space.
If then is invertible, moreover
3. Periodic Solutions for Equation (1)
Lemma 3.1 Let. If and. Then
Proof. Let. Then by applying the Fourier transform, we obtain that
Integration by parts we obtain that
The proof is complete.
Lemma 3.2 [1] Let and. Then the following assertions are equivalent:
1) and there exists such that
2) for any.
Let
By a Lemma 3.2 we obtain that
():such that and there exists with
Definition 3.1 [2] . For, we say that a sequence is an -multiplier, if for each there exists such that
Lemma 3.3 [2] Let and (is the set of all boun- ded linear operators from X to X). Then the following assertions are equivalent:
1) is an -multiplier.
2) is an -multiplier.
3.1. Existence of Strong Solutions for Equation (2)
Let.
Then the Equation (1) is equivalent:
Denote by; and for all. We define
We begin by establishing our concept of strong solution for Equation (2).
Definition 3.2 Let. A function is said to be a 2p- periodic strong -solution of Equation (2) if for all and Equation (2) holds almost every where.
Lemma 3.4 Let be a bounded linear operateur. Then
Proof. Let. Then
Moreover
It follows
Since G is bounded, then
Then
Lemma 3.5 [1] Let X be a Banach space, independent, symmetric, -valued random variables on a probability space, and such that, for each. Then
Proposition 3.1 Let A be a closed linear operator defined on an UMD space X. Suppose that. Then the following assertions are equivalent:
1) is an -multiplier for
2) is R-bounded.
Proof. 1) Þ 2) As a consequence of Proposition 2.1
2) Þ 1) We claim first that the set is R-bounded. In fact, for we have:
Since
Then
By Lemma 3.4, we obtain that
We conclude that
.
Next define, where. By Theorem 2.1 it is su- fficient to prove that the set is R-bounded. Since
we have
Therefore
Since products and sums of R-bounded sequences is R-bounded [10. Remark 2.2]. Then the proof is complete.
Lemma 3.6 Let. Suppose that and that for every
there exists a 2p-periodic strong -solution x of Equation (2). Then, x is the unique 2p-periodic strong -solution.
Proof. Suppose that and two strong -solution of Equation (2) then
is a strong -solution of Equation (2) corresponding to. Taking Fourier transform in (2), we obtain that
Then
It follows that for every and therefore. Then.
Theorem 3.1 Let X be a Banach space. Suppose that for every there exists a unique strong solution of Equation (2) for. Then
1) for every the operator has bounded inverse
2) is R-bounded.
Before to give the proof of Theorem 3.1, we need the following Lemma.
Lemma 3.7 if for all, then is a 2p-periodic strong -solution of the following equation
Proof of Lemma 3.7.
Then
We have and
Proof of Theorem 3.1: 1) Let and. Then for, there exists such that:
Taking Fourier transform, G and D are bounded. We have
by Lemma 3.2 and Lemma 3.4 , we deduce that:
Consequently, we have
is surjective.
If, then by Lemma 3.7, is a 2p-periodic strong -solution of Equation (2) corresponing to the function Hence and then is injective.
2) Let. By hypothesis, there exists a unique such that the Equation (2) is valid. Taking Fourier transforms, we deduce that
Hence
Since then there exists such that
Then is an -multiplier and is R-bounded.
3.2. Periodic Mild Solutions of Equation (2) When A Generates a C<sub>0</sub>-Semigroup
It is well known that in many important applications the operator A can be the infini- tesimal generator of -semigroup on the space X.
Definition 3.3 Assume that A generates a -semigroup on X. A func- tion x is called a mild solution of Equation (2) if:
Remark 3.1 ( [3] , Remark 4.2) Let be the -semigroup generated by A.
If is a continuous function, then and
Lemma 3.8 [3] Assume that A generates a -semigroup on X, if x is a mild solution then
Theorem 3.2 Assume that A generates a -semigroup on X and
. For some; if x is a mild solution of Equation (2). Then
Proof. Let x be a mild solution of Equation (2). Then by Lemma 3.8, we have
For, we have
Since:, then
which shows that the assertion holds for.
Now, define and by Lemma 3.1 We have:
Then
Corollary 3.1 Assume that A generates a -semigroup on X and let and x be a mild solution of Equation (2). If
has a bounded inverse. Then
Proof. From Theorem (3.2), we have that
Our main result in this work is to establish that the converse of Theorem 3.1 and Corollary 3.1 are true, provided X is an UMD space.
Theorem 3.3 Let X be an UMD space and be an closed linear operator. Then the following assertions are equivalent for.
1) for every there exists a unique 2p-periodic strong -solution of Equation (2).
2) and is R-bounded.
Lemma 3.9 [1] Let. If and for all Then
Proof of Theorem 3.3:
1) Þ 2) see Theorem 3.1
1) Ü 2) Let. Define.
By proposition 3.1, the family is an -multiplier it is equivalent to the family is an -multiplier that maps into, namely
there exists such that
In particular, and there exists such that
By Theorem 2.2, we have
Hence in, we obtain that
Since G is bounded, then
Using now (3) and (4) we have:
Since A is closed, then [Lemma 4.1] and from the uniqueness theorem of Fourier coefficients, that Equation (2) is valid.
Theorem 3.4 Let. Assume that A generates a -semigroup on X. If and is an -multiplier Then there exists a unique mild periodic solution of Equation (2).
Proof. For, we define
By Theorem 2.2 we can assert that as for the norm in.
We have is an -multiplier then there exists such that
let
Using again Theorem 2.2, we obtain that and is strong - solution of Equation (2) and verified
let. Then
For, we obtain that
From which we infer that the sequence is convergent to some element y as
. Moreover, y satisfies the following condition
let n go to infinity in (5), we can write
Then, we conclude that x is a 2p-periodic mild solution of Equation (2).
4. Applications
Example 5.1: Let A be a closed linear operator on a Hilbert space H and suppose that
and.
If then for every there exists a unique strong -
solution of Equation (2).
From the identity
it follows that is invertible whenever [Theo-
rem 2.3], we observe that.
Hence,
Then and by Theorem 2.3 we deduce that
Moreovery
and
We conclude that there exists a unique strong -solution of Equation (2). Using Corollary 3.8 in [2] .
Example 5.2:
Let A be a closed linear operator and X be a Hilbert space such that and. Suppose that. Then using Lemma 2.1
(1), we obtain that
From the identity it follows that is invertible whenever
Observe that.
Hence
Then and by Theorem 2.3, we have
Finaly
This proves that is R-bounded and by Theorem 3.3, we get that there exists a unique strong -solution of (2).
Acknowledgements
The authors would like to thank the referee for his remarks to improve the original version.
Cite this paper
Bahloul, R., Ezzinbi, K. and Sidki, O. (2016) Periodic Solutions in UMD Spaces for Some Neutral Partial Functional Differential Equations. Advances in Pure Mathematics, 6, 713-726. http://dx.doi.org/10.4236/apm.2016.610058
ReferencesArend, W. and Bu, S. (2002) The Operator-Valued Marcinkiewicz Multiplier Theorem and Maximal Regularity. Mathematische Zeitschrift, 240, 311-343. http://dx.doi.org/10.1007/s002090100384Lizama, C. (2006) Fourier Multipliers and Perodic Solutions of Delay Equatons in Banach Spaces. Journal of Mathematical Analysis and Applications, 324, 921-933. http://dx.doi.org/10.1016/j.jmaa.2005.12.043Henriquez, H.R., Pierri, M. and Prokopczyk, A. (2012) Periodic Solutions of Abstract Neutral Functional Differential Equations. Journal of Mathematical Analysis and Applications, 385, 608-621. http://dx.doi.org/10.1016/j.jmaa.2011.06.078Benkhalti, R., Bouzahir, H. and Ezzinbi, K. (2001) Existence of a Periodic Solution for Some Partial Functional Differential Equations with Infinite Delay. Journal of Mathematical Analysis and Applications, 256, 257-280. http://dx.doi.org/10.1006/jmaa.2000.7321Adimy, M., Bouzahir, H. and Ezzinbi, K. (2004) Existence and Stability for Some Partial Nentral Functional Differential Equations with Infinite Delay. Journal of Mathematical Analysis and Applications, 294, 438-461. http://dx.doi.org/10.1016/j.jmaa.2004.02.033Adimy, M., Ezzinbi, K. and Laklach, M. (2000) Local Existence and Global Continuation for a Class of Partial Neutral Functional Differential Equations. Comptes Rendus de l’Académie des Sciences—Series I—Mathematics, 330, 957-962.Adimy, M. and Ezzinbi, K. (1998) A Class of Linear Partial Neutral Functional Differential Equations with Non-Dense Domain. Journal of Differential Equations, 147, 285-332. http://dx.doi.org/10.1006/jdeq.1998.3446Adimy, M and Ezzinbi, K. (1998) Local Existence and Linearized Stability for Partial Functional Differential Equations. Dynamic Systems and Applications, 7, 389-404.Bourgain, J. (1983) Some Remarks on Banach Spaces in Which Martingale Differences Sequences Are Unconditional. Arkiv för Matematik, 21, 163-168. http://dx.doi.org/10.1007/BF02384306