메인 Abstract and Applied Analysis Strongly nonlinear potential theory on metric spaces

Strongly nonlinear potential theory on metric spaces

이 책이 얼마나 마음에 드셨습니까?
파일의 품질이 어떻습니까?
책의 품질을 평가하시려면 책을 다운로드하시기 바랍니다
다운로드된 파일들의 품질이 어떻습니까?
권:
7
년:
2002
언어:
english
기록:
Abstract and Applied Analysis
DOI:
10.1155/s1085337502203024
파일:
PDF, 1.91 MB
Conversion to is in progress
Conversion to is failed
0 comments
 

리뷰를 게시하기에 로그인 또는 가입 하세요
책에 대한 리뷰를 작성하거나 귀하의 독서 경험을 공유할 수 있습니다. 다른 독자들이 귀하가 읽은 책에 대한 의견에 귀를 기울일 것입니다. 개인적으로 책이 마음에 들었거나 그렇지 않았거나 정직하고 상세한 호평은 다른 독자들이 자신에게 적합한 책을 찾는데 도움이 됩니다.
1

Bounded solutions of nonlinear Cauchy problems

년:
2002
언어:
english
파일:
PDF, 1.92 MB
2

Generic uniqueness of a minimal solution for variational problems on a torus

년:
2002
언어:
english
파일:
PDF, 1.87 MB
STRONGLY NONLINEAR POTENTIAL
THEORY ON METRIC SPACES
NOUREDDINE AÏSSAOUI
Received 23 January 2002

We define Orlicz-Sobolev spaces on an arbitrary metric space with a Borel regular outer measure, and we develop a capacity theory based on these spaces. We
study basic properties of capacity and several convergence results. We prove that
each Orlicz-Sobolev function has a quasi-continuous representative. We give estimates for the capacity of balls when the measure is doubling. Under additional
regularity assumption on the measure, we establish some relations between capacity and Hausdorff measures.
1. Introduction
The introduction and the extensive study of Sobolev spaces on arbitrary metric
spaces by Franchi et al. [9], Hajłasz [12], Hajłasz and Koskela [13], Hajłasz and
Martio [14], and others, have given a great impulse to several developments in
geometric analysis on metric measure spaces. Important examples are the substantial progress of various domains such as fractals, partial differential equations, Carnot-Carathéodory geometries, stochastic process, and so forth.
The nonlinear potential theory on metric spaces has seen a great jump since
the development of the capacity theory in these spaces by Kilpeläinen et al. [16],
Kinnunen and Martio [17], and others.
For Orlicz and Orlicz-Sobolev spaces in the Euclidean space, we have developed in [2, 3, 4, 5, 6, 7] a potential theory, called strongly nonlinear potential
theory. It is natural to develop this theory in the setting of metric spaces. It is the
object of this paper.
A Lipschitz characterization of Orlicz-Sobolev spaces in Euclidean case is
given. Since a density argument and the Hardy-Littlewood maximal function
are involved, we must suppose the Orlicz space to be reflexive. This characterization is used to introduce a definition of Orlicz-Sobolev spaces on an arbitrary
metric measure space. We prove an approximation theorem of Orlicz-Sobolev
Copyright © 2002 Hindawi Publishing Corporation
Abstract and Applied Analysis 7:7 (2002) 357–374
2000 Math; ematics Subject Classification: 46E35, 31B15, 28A80
URL: http://dx.doi.org/10.1155/S1085337502203024

358

Strongly nonlinear potential theory on metric spaces

functions by Lipschitz functions, both in Lusin and in norm sense. This generalizes a result by Hajłasz in [12] relative to the Sobolev case.
Then we develop a capacity theory based on the definition of Orlicz-Sobolev
spaces on an arbitrary metric measure space. Basic properties of capacity and
several convergence results are studied. Moreover, we prove that each OrliczSobolev function has a quasi-continuous representative and we give estimates
for the capacity of balls when the measure is doubling. Some relations between
capacity and Hausdorff measures are established under additional regularity assumption on the measure.
This paper is organized as follows. In Section 2, we list the prerequisites from
the Orlicz theory. Section 3 is dedicated to Lipschitz characterization of OrliczSobolev spaces in the Euclidean case, to the study of Orlicz-Sobolev spaces on
metric spaces and to establish an approximation theorem of Orlicz-Sobolev
functions by Lipschitz functions. Section 4 is reserved to establish important
properties of capacity on metric spaces. Section 5 deals with the comparison between capacity and measure in the metric space, and between capacity and the
Hausdorff measure.
2. Preliminaries
An N-function is a continuous convex and even function Φ defined on R, verifying Φ(t) > 0 for t > 0, limt→0 Φ(t)/t = 0, and limt→+∞ Φ(t)/t = +∞.
 |t|
We have the representation Φ(t) = 0 ϕ(x)dL(x), where ϕ : R+ → R+ is nondecreasing, right continuous, with ϕ(0) = 0, ϕ(t) > 0 for t > 0, limt→0+ ϕ(t) = 0,
and limt→+∞ ϕ(t) = +∞. Here L stands for the Lebesgue measure. We put in the
sequel, as usual, dx = dL(x).
 |t|
The N-function Φ∗ conjugate to Φ is defined by Φ∗ (t) = 0 ϕ∗ (x)dx, where
ϕ∗ is given by ϕ∗ (s) = sup{t : ϕ(t) ≤ s}.
Let Φ be an N-function. We say that Φ verifies the ∆2 condition if there is a
constant C > 0 such that Φ(2t) ≤ CΦ(t) for all t ≥ 0. We denote by C(Φ) the
smallest of such constants. That is, C(Φ) = supt>0 Φ(2t)/Φ(t).
The ∆2 condition for Φ can be formulated in the following equivalent way:
for every C > 0 there exists C  > 0 such that Φ(Ct) ≤ C  Φ(t) for all t ≥ 0.
Let (X,Γ,µ) be a measure space and Φ an N-function. The Orlicz class ᏸΦ,µ (X)
is defined by




ᏸΦ,µ (X) = f : X −→ R measurable :



X





Φ f (x) dµ(x) < ∞ .

(2.1)

We define the Orlicz space LΦ,µ (X) by


LΦ,µ (X) = f : X −→ R measurable :





X





Φ α f (x) dµ(x) < ∞ for some α > 0 .
(2.2)

Noureddine Aı̈ssaoui 359
We always have ᏸΦ,µ (X) ⊂ LΦ,µ (X). The equality ᏸΦ,µ (X) = LΦ,µ (X) occurs if
Φ verifies the ∆2 condition.
The Orlicz space LΦ,µ (X) is a Banach space with the following norm, called
the Luxemburg norm,
 
| f |

Φ,µ,X





f (x)
= inf r > 0 :
Φ
dµ(x) ≤ 1 .

r

X

(2.3)

the ∆2 condition.
Recall that LΦ,µ (X) is reflexive if Φ and Φ∗ verify

Note that if Φ verifies the ∆2 condition, then Φ( fi (x))dµ → 0 as i → ∞ if and
only if | fi |Φ,µ,X → 0 as i → ∞.
Let Ω be an open set in RN , C∞ (Ω) be the space of functions which, together with all their partial derivatives of any order, are continuous on Ω, and
∞
N
∞
N
C∞
0 (R ) = C0 stands for all functions in C (R ) which have compact support
N
k
in R . The space C (Ω) stands for the space of functions having all derivatives
of order ≤ k continuous on Ω, and C(Ω) is the space of continuous functions
on Ω.
The (weak) partial derivative of f of order |β| is denoted by
Dβ f =

∂|β|
β

β

β

∂x1 1 · ∂x2 2 · · · · · ∂xNN

f.

(2.4)

Let Φ be an N-function and m ∈ N. We say that a function f : RN → R has a
distributional (weak partial) derivative of order m, denoted by Dβ f , |β| = m, if


f Dβ θ dx = (−1)|β|







Dβ f θ dx,

∀θ ∈ C∞
0 .

(2.5)

Let Ω be an open set in RN and denote LΦ,L (Ω) by LΦ (Ω). The Orlicz-Sobolev
space W m LΦ (Ω) is the space of real functions f , such that f and its distributional
derivatives up to the order m, are in LΦ (Ω).
The space W m LΦ (Ω) is a Banach space equipped with the norm
 
| f |

m,Φ

=
0≤|β|≤m

 β 
 D f  ,
Φ

f ∈ W m LΦ (Ω),

(2.6)

where |Dβ f |Φ = |Dβ f |Φ,L,Ω .
Recall that if Φ verifies the ∆2 condition, then C∞ (Ω) ∩ W m LΦ (Ω) is dense in
m
N
m
N
W LΦ (Ω), and C∞
0 (R ) is dense in W LΦ (R ).
For more details on the theory of Orlicz spaces, see [1, 18, 19, 20, 21].
3. Orlicz-Sobolev spaces on metric spaces
3.1. The Euclidean case. We begin by two lemmas which lead to a Lipschitz
characterization of Orlicz-Sobolev spaces.

360

Strongly nonlinear potential theory on metric spaces

Lemma 3.1. Let Φ be an N-function satisfying the ∆2 condition, Q a cube in RN ,
and f ∈ W 1 LΦ (Q). Then, for almost all x ∈ Q,


f (x) − fQ ≤ C
where fQ = 1/µ(Q)


Q

Q

∇ f (y)
d y,
|x − y |N −1

(3.1)

f dµ, and the constant C depends only on N and Q.

Proof. It is enough to establish (3.1) for C1 (Q). But in this case the proof can be

found in [11, Lemma 7.16].
We omit the proof of the following lemma (Hedberg’s inequality) since it is
exactly the same as the one in [15] or in [23, Lemma 2.8.3].
Lemma 3.2. Let Φ be an N-function and f ∈ W 1 LΦ (RN ). Then there is a constant
C depending only on N such that for all x ∈ RN ,

B(x,R)



∇ f (y)
d y ≤ CRᏹR |∇ f | (x),
|x − y |N −1

where ᏹR (h)(x) = supr<R (1/ |B(x,r)|)
measure of the ball B(x,r) on RN .



B(x,r) |h(y)| d y

(3.2)

and |B(x,r)| is the Lebesgue

Now we give a Lipschitz characterization of Orlicz-Sobolev spaces.
Theorem 3.3. Let Φ be an N-function such that Φ and Φ∗ satisfy the ∆2 condition. Then f ∈ W 1 LΦ (RN ) if and only if there exists a nonnegative function
g ∈ LΦ (RN ) such that
f (x) − f (y) ≤ |x − y | g(x) + g(y) ,

(3.3)

for all x, y ∈ RN \ F, |F | = 0.
Proof. Let Q be a cube in RN and x, y ∈ Q. Then we can find a subcube Q∗ with
x, y ∈ Q∗ and diamQ∗ ≈ |x − y |. Here A ≈ B if there is a constant C such that
C −1 ≤ A ≤ CB. Let f ∈ W 1 LΦ (RN ). Then by (3.1) and (3.2) we get
f (x) − f (y) ≤ f (x) − fQ∗ + f (y) − fQ∗





 
≤ C |x − y | ᏹ|x− y | |∇ f | (x) + ᏹ|x− y | |∇ f | (y)
 

 

≤ C |x − y | ᏹ |∇ f | (x) + ᏹ |∇ f | (y) ,

(3.4)



where ᏹ(h)(x) = sup0<r (1/ |B(x,r)|) B(x,r) |h(y)| d y is the Hardy-Littlewood
maximal function. Since Φ∗ verifies the ∆2 condition, by [10] the maximal operator is bounded in LΦ (RN ). Hence, there is a nonnegative function g ∈ LΦ (RN )
such that
f (x) − f (y) ≤ |x − y | g(x) + g(y)

a.e.

(3.5)

Noureddine Aı̈ssaoui 361
For the reverse implication, it suffices to show that, due to Riesz representation and Radon-Nikodym theorem, there is a nonnegative function h ∈ LΦ (RN )
such that
∂f
def
[φ] = −
∂xi





f

∂φ
dx ≤ |φ|hdx
∂xi

(3.6)

for all φ ∈ C∞
0 . Now, integrating (3.3) twice over a ball B(x,ε), we get
1
B(x,ε)


B(x,ε)

f (x) − fQ∗ dx ≤ Cε

1
B(x,ε)


B(x,ε)

g(x)dx.

(3.7)



−N
ψ(x/ε). We
Let ψ ∈ C∞
0 (B(0,1)) be such that ψ(x)dx = 1. Put ψε (x) = ε
have





f





∂φ
∂φ 
∂ψε
dx = lim
ψε ∗ f (x)dx = − lim φ(x)
∗ f (x)dx.
ε→0
ε→0
∂xi
∂xi
∂xi

(3.8)



Since (∂ψε /∂xi )(x)dx = 0, we get

 ∂ψε
∂ψε
∗ f = f − fB(x,ε) ∗
.
∂xi
∂xi

(3.9)

Hence


∂ψε
∗ f (x) ≤ Cε−N −1
f − fB(x,ε) (x)dx
∂xi
B(x,ε)

1
≤C
g(x)dx.
B(x,ε) B(x,ε)

(3.10)

Thus




∂φ
f
dx ≤ |φ|ᏹ(g)(x)dx.
∂xi

(3.11)

Now by [10], ᏹ(g) ∈ LΦ (RN ) since Φ∗ satisfies the ∆2 condition. The proof

is complete.
3.2. The metric case. Let (X,d) be a metric space and let µ be a nonnegative
Borel regular outer measure on X. The triplet (X,d,µ) will be fixed in the sequel
and will be denoted by X.
Let u : X → [−∞,+∞] be a µ-measurable function defined on X. We denote
by D(u) the set of all µ-measurable functions g : X → [0,+∞] such that




u(x) − u(y) ≤ d(x, y) g(x) + g(y)

(3.12)

for every x, y ∈ X \ F, x = y, with µ(F) = 0. The set F is called the exceptional set
for g.

362

Strongly nonlinear potential theory on metric spaces

Note that the right-hand side of (3.12) is always defined for x = y. For the
points x, y ∈ X, x = y such that the left-hand side of (3.12) is undefined, we may
assume that the left-hand side is +∞.
Let Φ be an N-function. The Dirichlet-Orlicz space L1Φ,µ (X) is the space of all
µ-measurable functions u such that D(u) ∩ LΦ,µ (X) = ∅. This space is equipped
with the seminorm
 
|u|

L1Φ,µ (X)

 

= inf |g |Φ,µ,X : g ∈ D(u) ∩ LΦ,µ (X) .

(3.13)

1
(X) = LΦ,µ (X) ∩ L1Φ,µ (X) equipped with the
The Orlicz-Sobolev space is MΦ,µ
norm

 
|u|

1
MΦ,µ
(X)

 
 
= |u|Φ,µ,X + |u|L1Φ,µ (X) .

(3.14)

Lemma 3.4. Let g1 ∈ D(u1 ), g2 ∈ D(u2 ), and α,β ∈ R. If g ≥ |α|g1 + |β|g2 almost
everywhere, then g ∈ D(αu1 + βu2 ).


Proof. The proof is a simple verification.

Lemma 3.5. Let Φ be an N-function. Let (ui )i and (gi )i be two sequences of functions such that, for all i,gi ∈ D(ui ). If ui → u in LΦ,µ (X) and gi → g in LΦ,µ (X),
then g ∈ D(u).
Proof. There are two subsequences of (ui )i and (gi )i , which we denote again by
(ui )i and (gi )i , such that ui → u and gi → g, µ-a.e. Let, for i = 1,2,..., Fi be the
G be a set of measure zero such that ui → u and gi →
exceptional set for gi and let

g on c G. We set F = G ∪ ( ∞
i=1 Fi ). It is clear that µ(F) = 0 and |u(x) − u(y)| ≤
d(x, y)(g(x) + g(y)), for all x, y ∈ X \ F. This implies that g ∈ D(u). The proof is

complete.
1
Theorem 3.6. Let Φ be an N-function. Then MΦ,µ
(X), equipped with the norm
defined by (3.14), is a Banach space.
1
1
(X) is a vector space. It remains to prove that MΦ,µ
(X)
Proof. It is clear that MΦ,µ
is complete for the norm defined by (3.14). Let (ui )i be an arbitrary Cauchy
1
(X). Taking if necessary a subsequence, we may assume that
sequence in MΦ,µ
−i
1
|(ui − ui+1 )|MΦ,µ
(X) ≤ 2 . Set v j = u j − u j+1 . There exists h j ∈ D(v j ) ∩ LΦ,µ (X)
−i
1
such that |h j |MΦ,µ
(X) ≤ 2 . Let g1 ∈ D(u1 ) be arbitrary and set, for k ≥ 2, gk =





g1 + kj =−11 h j . From the identity uk = u1 − kj =−11 v j and Lemma 3.4, it follows that
gk ∈ D(u), for all k ∈ N. Now (uk )k and (gk )k are Cauchy sequences in LΦ,µ (X).
Thus there are limit functions u = limuk in LΦ,µ (X) and g = limgk in LΦ,µ (X).
Lemma 3.5 implies that g ∈ D(u) and therefore the sequence (uk )k has a limit in
1
(X). The proof is complete.

MΦ,µ
1
The previous results lead to the following characterization of MΦ,µ
(X).

Noureddine Aı̈ssaoui 363
1
(X) if and only if u ∈
Lemma 3.7. Let Φ be an N-function. The function u ∈ MΦ,µ
LΦ,µ (X) and there are functions ui ∈ LΦ,µ (X), i = 1,2,..., such that ui → u µ-a.e.
and gi ∈ D(ui ) ∩ LΦ,µ (X), i = 1,2,..., such that gi → g µ-a.e. for some g ∈ LΦ,µ (X).

Proof. The proof is an immediate consequence of Lemma 3.5 and Theorem 3.6.

1
Moreover, MΦ,µ
(X) satisfies the following lattice property.
1
Lemma 3.8. Let Φ be an N-function. Let u1 ,u2 ∈ MΦ,µ
(X). If g1 ∈ D(u1 ) and
g2 ∈ D(u2 ), then
1
(i) u = max(u1 ,u2 ) ∈ MΦ,µ
(X) and max(g1 ,g2 ) ∈ D(u) ∩ LΦ,µ (X);
1
(ii) v = min(u1 ,u2 ) ∈ MΦ,µ (X) and max(g1 ,g2 ) ∈ D(v) ∩ LΦ,µ (X).

Proof. We prove the case (i) only, the proof of (ii) is similar. Let g = max(g1 ,g2 )
and suppose that F1 and F2 are the exceptional sets for u1 and u2 in (3.12),
respectively. It is evident that u,g ∈ LΦ,µ (X). It remains to show that g ∈ D(u).
Let U = {x ∈ X (F1 ∪ F2 ) : u1 (x) ≥ u2 (x)}. Let x, y ∈ U. Then




u(x) − u(y) = u1 (x) − u1 (y) ≤ d(x, y) g1 (x) + g1 (y) .

(3.15)

By the same manner we obtain, for x, y ∈ X U,




u(x) − u(y) ≤ d(x, y) g2 (x) + g2 (y) .

(3.16)

For the remaining cases, let x ∈ U and y ∈ X U. If u1 (x) ≥ u2 (y), then




u(x) − u(y) = u1 (x) − u2 (y) ≤ u1 (x) − u1 (y) ≤ d(x, y) g1 (x) + g1 (y) .
(3.17)
If u1 (x) < u2 (y), then




u(x) − u(y) = −u1 (x) + u2 (y) ≤ −u2 (x) + u2 (y) ≤ d(x, y) g2 (x) + g2 (y) .
(3.18)
The case x ∈ X U and y ∈ U follows by symmetry. Hence, for all x, y ∈
X (F1 ∪ F2 ) with µ(F1 ∪ F2 ) = 0, |u(x) − u(y)| ≤ d(x, y)(g(x) + g(y)).

Next, we prove the following important Poincaré inequality for OrliczSobolev functions.
1
(X) and E ⊂ X is µ-measuProposition 3.9. Let Φ be an N-function. If u ∈ MΦ,µ
rable with 0 < µ(E) < ∞, then for every g ∈ D(u) ∩ LΦ,µ (X),



 
 u − uE  1
 
LΦ,µ (E) ≤ 2diam(E) |g | L1Φ,µ (E) ,

where uE = 1/µ(E)


E

f dµ.

(3.19)

364

Strongly nonlinear potential theory on metric spaces

Proof. Put C = 2diam(E)|g |L1Φ,µ (E) and remark that




1
µ(E)

u − uE (x) =







u(x) − u(y) dµ(y).

E

(3.20)

By the Jensen inequality

E

Φ



u − uE
(x)dµ(x) =
C




E

Φ

 

≤

1
µ(E)





u(x) − u(y)
dµ(y) (x)dµ(x)
C
E




u(x) − u(y)
1
dµ(y) dµ(x).
Φ
µ(E)
C
E

E

(3.21)

On the other hand, using the definition of D(u) and the convexity of Φ, we get




u(x) − u(y)
g(x) − g(y)

Φ
≤Φ 
C
2|g |L1Φ,µ (E)
 






g(x)
g(y)
1


Φ 
+Φ 
≤
|g | 1
|g | 1
2
LΦ,µ (E)
LΦ,µ (E)
Now we use the fact that

E

Φ



E Φ(g(x)/ |g |L1Φ,µ (E) )dµ(x) ≤ 1



(3.22)

.

to deduce that



u − uE
(x)dµ(x)
C



   
g(x)
g(y)
1
 

≤
Φ 
+
Φ
dµ(y)dµ(x) (3.23)
|g | 1
|g | 1
2µ(E) E E
LΦ,µ (E)
LΦ,µ (E)
1
≤
2µ(E)

 
E



g(x)


1+Φ 
|g |

The proof is complete.

L1Φ,µ (E)





µ(E) dµ(x) ≤ 1.


Now, we prove an approximation theorem of Orlicz-Sobolev functions by
Lipschitz functions, both in Lusin sense and in norm. This generalizes a result in
[12] relative to the Sobolev case.
Theorem 3.10. Let Φ be an N-function satisfying the ∆2 condition and u ∈
1
MΦ,µ
(X). Then for every ε > 0, there is a Lipschitz function h such that
(1) µ({x : u(x) = h(x)}) < ε;
1
(2) |u − h|MΦ,µ
(X) < ε.
1
Proof. Let u ∈ MΦ,µ
(X) and let g be taken from the definition of |u|L1Φ,µ (X) . Let
Xn = {x ∈ X : |u(x)| ≤ n and g(x) ≤ n}. Since u,g ∈ LΦ,µ(X) and Φ satisfies the ∆2
condition, we conclude that u,g ∈ ᏸΦµ (X). This implies that limn→∞ Φ(n)µ(c Xn )
= 0, and limn→∞ µ(c Xn ) = 0 because Φ(n) → ∞ when n → ∞. The restriction u|Xn

Noureddine Aı̈ssaoui 365
is Lipschitz with the constant 2n and can be extended to the Lipschitz function
u on X with the same constant (see [8, Section 2.10.4] and [22, Theorem 5.1]).
We put un = (sgn u )min(|u |,n). It is clear that un is Lipschitz with the constant
n→∞
2n, un |Xn = u|Xn , |un | ≤ n and µ({x : u(x) = un (x)}) ≤ µ(c Xn ) −−−→ 0.
1
We must simply prove that un → u in MΦ,µ (X) when n → ∞. We have








Φ u − un dµ =

≤



cX

cX



Φ u − un dµ
n



Φ |u| + un
n





dµ





 

CΦ
Φ |u| dµ +
Φ un dµ
cX
cX
2
n
n




c  n→∞
CΦ
Φ |u| dµ + Φ(n)µ Xn −−−→ 0.
≤
cX
2
n

(3.24)

≤

n→∞

Since Φ satisfies the ∆2 condition, |u − un |L1Φ,µ (X) −−−→ 0. It remains to estimate the gradient. Let

0

for x ∈ Xn ,
gn = 
g(x) + 3n for x ∈c Xn .

(3.25)


We
 have |(u − un )(x) − (u − un )(y)| ≤ d(x, y)(gn (x) + gn (y)). Since Φ(gn )dµ =
c X Φ(gn )dµ → 0 as n → ∞, we conclude that |gn |L1 (X) → 0 when n → ∞ ben
Φ,µ
cause Φ satisfies the ∆2 condition. The theorem follows.

4. Capacity on metric spaces
Definition 4.1. For a set E ⊂ X, define CΦ,µ (E) by
 

CΦ,µ (E) = inf |u|MΦ,µ
1
(X) : u ∈ B(E) ,

(4.1)

1
(X) : u ≥ 1 on a neighborhood of E}.
where B(E) = {u ∈ MΦ,µ
If B(E) = ∅, we set CΦ,µ (E) = ∞. Functions belonging to B(E) are called admissible functions for E.

Remark 4.2. In the definition of CΦ,µ (E), we can restrict ourselves to those admissible functions u such that 0 ≤ u ≤ 1.
Proof. Let B  (E) = {u ∈ B(E) : 0 ≤ u ≤ 1}, then B (E) ⊂ B(E) implies that
 


CΦ,µ (E) ≤ inf |u|MΦ,µ
1
(X) : u ∈ B (E) .

(4.2)

366

Strongly nonlinear potential theory on metric spaces

1
On the other hand, let ε > 0 and take u ∈ B(E) such that |u|MΦ,µ
(X)

≤ CΦ (E) + ε. Then v = max(0,min(u,1)) ∈ B (E), and by Lemma 3.8 we get
D(u) ⊂ D(v). Hence,

 
 
  

  1
  1
inf |u|MΦ,µ
1
(X) : u ∈ B (E) ≤ |v | MΦ,µ
(X) ≤ |u| MΦ,µ
(X) ≤ CΦ,µ (E) + ε.

(4.3)


This completes the proof.

We define a capacity as an increasing positive set function C given on a σadditive
class 
of sets Γ, which contains compact sets and such that C(∅) = 0 and

C( i≥1 Xi ) ≤ i≥1 C(Xi ) for Xi ∈ Γ, i = 1,2,.... C is called outer capacity if for
every X ∈ Γ,




C(X) = inf C(O) : O open, X ⊂ O .

(4.4)

Theorem 4.3. Let Φ be an N-function. The set function CΦ,µ is an outer capacity.
Proof. It is obvious that CΦ,µ (∅) = 0 and that CΦ,µ is increasing. For countable
subadditivity,
let Ei , i = 1,2,... be subsets of X and let ε > 0. We may assume that
∞
C
(E
)
<
∞. We choose ui ∈ B(Ei ) and gui ∈ D(ui ) ∩ LΦ,µ (X) so that for
i=1 Φ,µ i
i = 1,2,...,




 
−i
 ui 
 gu 
i
Φ,µ,X +
Φ,µ,X ≤ CΦ Ei + ε2 .

(4.5)



We show that v = supi ui is admissible for ∞
i=1 Ei and g = supi gui ∈ D(v) ∩
LΦ,µ (X).
Observe that v,g ∈ LΦ,µ (X). Define vk = max1≤i≤k ui . By Lemma 3.8 the function gvk = max1≤i≤k gui ∈ D(vk ) ∩ LΦ,µ (X). Since vk → v µ-a.e., Lemma 3.7 gives

∞
1
v ∈ MΦ,µ
(X). Clearly v ≥ 1 in a neighborhood of ∞
i=1 Ei . Hence CΦ,µ ( i=1 Ei ) ≤
1
|v |MΦ,µ
(X) .
By [6, Lemma 2],
 
|v |

∞
1
MΦ,µ
(X)

≤
i=1






 gu 
 ui 
+
i
Φ,µ,X
Φ,µ,X ≤


Since ε is arbitrary, we deduce that CΦ,µ ( ∞
i=1 Ei ) ≤
is a capacity.
It remains to prove that CΦ,µ is outer, that is,


∞

 

CΦ,µ Ei + ε.

(4.6)

i=1

∞

i=1 CΦ,µ (Ei ).



CΦ,µ (E) = inf CΦ,µ (O) : E ⊂ O, O open .

Hence CΦ,µ

(4.7)

By monotonicity, CΦ,µ (E) ≤ inf {CΦ,µ (O) : E ⊂ O, O open}. For the reverse in1
equality, let ε > 0 and let u ∈ B(E) be such that |u|MΦ,µ
(X) ≤ CΦ,µ (E) + ε. Since

Noureddine Aı̈ssaoui 367
u ∈ B(E), there is an open set O, E ⊂ O, such that u ≥ 1 on O. This implies that
1
CΦ,µ (O) ≤ |u|MΦ,µ
(X) ≤ CΦ,µ (E) + ε.
Since ε is arbitrary, we obtain the claim, and the proof is complete.

Corollary 4.4. Let Φ be an 
N-function. Let (Ki )i≥1 be a decreasing sequence of
compact sets in X and let K = ∞
i=1 Ki . Then
 

CΦ,µ (K) = lim CΦ,µ Ki .
i→∞

(4.8)

Proof. This is a direct consequence of the fact that CΦ,µ is an outer capacity. 
Theorem 4.5. Let Φ be a uniformly convex N-function such that Φ satisfies the
∆ condition. Let (Oi )i≥1 be an increasing sequence of open sets in X and let O =
2∞
i=1 Oi . Then




CΦ,µ (O) = lim CΦ,µ Oi .
i→∞

(4.9)

Proof. The hypothesis implies that LΦ,µ (X) is uniformly convex. By monotonicity, limi→∞ CΦ,µ (Oi ) ≤ CΦ,µ (O). To prove the reverse inequality, we may assume
that limi→∞ CΦ,µ (Oi ) < ∞. Let ε > 0 and for i = 1,2,..., ui ∈ B(Oi ), and gui ∈
D(ui ) ∩ LΦ,µ (X) be such that




 
 ui 
 gu 
i
Φ,µ,X +
Φ,µ,X ≤ CΦ,µ Oi + ε.

(4.10)

The sequence (ui )i is bounded in LΦ,µ (X) and, hence, it possesses a weakly
convergent subsequence, which we denote again by (ui )i . The sequence (gui )i is
also bounded in LΦ,µ (X) and, by passing to a subsequence, we may assume that
ui → u weakly in LΦ,µ (X) and gui → g weakly in LΦ,µ (X). We use the Banach-Saks
j
theorem to deduce that the sequence defined by v j = j −1 i=1 ui converges to u
j
in LΦ,µ (X) and gv j = j −1 i=1 gui converges to g in LΦ,µ (X).
Now there is a subsequence of the sequence (v j ) j so that v j → u µ-a.e. and
1
gv j → g µ-a.e. By Lemma 3.7, u ∈ MΦ,µ
(X). On the other hand, v j → 1 µ-a.e. in O
and hence u ≥ 1 µ-a.e. in O. Thus u ∈ B(O). By the weak lower semicontinuity
of norms, we get








CΦ,µ (O) ≤ |u|Φ,µ,X + |g |Φ,µ,X






≤ liminf  ui Φ,µ,X +  gui Φ,µ,X
i→∞
 
≤ limCΦ,µ Oi + ε.

(4.11)

i→∞

By letting ε → 0, we obtain the result.



The set function CΦ,µ is called the Φ-capacity. If a statement holds except
on a set E where CΦ,µ (E) = 0, then we say that the statement holds Φ-quasieverywhere (abbreviated Φ-q.e.).

368

Strongly nonlinear potential theory on metric spaces

1
(X). If PLu =
Lemma 4.6. Let Φ be any N-function and let u be a function in MΦ,µ
{x ∈ X : limt→x u(t) = ∞} denotes the set of poles of u, then CΦ,µ (PLu ) = 0.

Proof. If n is any positive integer, the function un = n−1 inf(u,n) is an admissi1
ble function for the Φ-capacity of the set PLu . It is easily seen that |un |MΦ,µ
(X) ≤
−1
1
1
n |u|MΦ,µ
(X) . This implies that limn→∞ |un |MΦ,µ
(X) = 0. Thus CΦ,µ (PLu ) = 0.

Theorem 4.7. Let Φ be an N-function and let E be any subset of X. Then
CΦ,µ (E) = 0 if and only if for any ε > 0, there exists a nonnegative function u ∈
1
1
MΦ,µ
(X) such that limt→x u(t) = ∞ for any x ∈ E and |u|MΦ,µ
(X) ≤ ε.
Proof. Let E ⊂ X be such that CΦ,µ (E) = 0, and let ε > 0. Then, by the definition
of the Φ-capacity, there is a sequence of nonnegative functions, (un )n , such that
−n
1
un = 1 in some neighborhood of E and |un |MΦ,µ
(X) ≤ 2 ε, for any n. Then the

1
function u = n un belongs to MΦ,µ
(X) and limt→x u(t) = ∞. Furthermore, it is
1
evident that |u|MΦ,µ
(X) ≤ ε.
The converse implication is an immediate consequence of Lemma 4.6.

Theorem 4.8. Let Φ be an N-function and (ui )i be a Cauchy sequence of func1
(X) ∩ C(X). Then, there is a subsequence (ui )i of (ui )i which contions in MΦ,µ
verges pointwise Φ-q.e. in X. Moreover, the convergence is uniform outside a set of
arbitrary small Φ-capacity.
Proof. Since (ui )i is a Cauchy sequence, there is a subsequence (ui )i of (ui )i such
that
∞

i=1



2i  ui − ui+1 MΦ,µ
1
(X) < ∞.

(4.12)


We set Xi = {x ∈ X : |ui (x) − ui+1 (x)| > 2−i } for i = 1,2,..., and Y j = ∞
i= j Xi .
Since the functions ui are continuous by hypothesis, Xi and Y j are open. Then,
for all i, 2i (ui − ui+1 ) ∈ B(Xi ). Hence


 
CΦ,µ Xi ≤ 2i  ui − ui+1 MΦ,µ
1
(X) .

(4.13)

By subadditivity of CΦ,µ , we obtain




∞

CΦ,µ Y j ≤

 

∞

CΦ,µ Xi ≤
i= j

i= j



2i  ui − ui+1 MΦ,µ
1
(X) .

(4.14)

From the convergence of the sum (4.12) we conclude that


CΦ,µ 

∞

j =1







Y j  ≤ lim CΦ,µ Y j = 0.
j →∞

(4.15)

Noureddine Aı̈ssaoui 369
Thus (ui )i converges in X 
and all k > j
uj (x) − uk (x) ≤

k −1

∞

j =1 Y j .

Moreover, we have for any x ∈ X Y j

ui (x) − ui+1 (x) ≤

i= j

k −1

2−i ≤ 21− j .

(4.16)

i= j

This implies that (ui )i converges uniformly in X Y j .
This completes the proof.



Definition 4.9. Let Φ be an N-function. A function u : X → [−∞,+∞] is Φquasi-continuous in X if for every ε > 0, there is a set E such that CΦ,µ (E) < ε
and the restriction of u to X E is continuous. Since CΦ,µ is an outer capacity,
we may assume that E is open.
Theorem 4.10. Let Φ be an N-function satisfying the ∆2 condition and u ∈
1
1
MΦ,µ
(X). Then there is a function v ∈ MΦ,µ
(X) such that u = v µ-a.e. and v is
Φ-quasi-continuous in X. The function v is called a Φ-quasi-continuous representative of u.
1
(X) is a Banach space and by Theorem 3.10, C(X) ∩
Proof. We know that MΦ,µ
1
1
MΦ,µ (X) is a dense subspace of MΦ,µ
(X). Hence, completeness implies that
1
1
MΦ,µ (X) can be characterized as the completion of C(X) ∩ MΦ,µ
(X) in the norm
1
of MΦ,µ (X). Thus there are sequences of functions (ui )i ⊂ C(X) ∩ LΦ,µ (X) and
(gi )i ⊂ D(ui − u) such that ui → u and gi → 0 in LΦ,µ (X). Hence for a subsequence, which we denote again by (ui )i , ui → u µ-a.e. From Theorem 4.8, we
deduce that the limit function v of the sequence (ui )i is Φ-quasi-continuous in

X. This completes the proof.

5. Comparison between capacity and measures
5.1. Comparison between capacity and the measure µ. The sets of zero capacity are exceptional sets in the strongly nonlinear potential theory. We show in
the next lemma that sets of vanishing capacity are also of measure zero.
Lemma 5.1. Let Φ be any N-function. Then
(1) if E ⊂ X is such that CΦ,µ (E) = 0, then µ(E) = 0;
(2) if E ⊂ X is such that µ(E) = 0, then CΦ,µ (E) ≥ 1/Φ−1 (1/µ(E)).
Proof. (1) If n is a strictly positive integer, we can find un ∈ B(E) such that
1
|nun |Φ,µ,X ≤ |nun |MΦ,µ
(X) ≤ 1. Hence, there is an open set On such that E ⊂
On and un ≥ 1 in On . Then




E



Φ nun dµ ≤



On





Φ nun dµ ≤





X



Φ nun dµ ≤ 1.

(5.1)

By the inequality nΦ(un ) ≤ Φ(nun ) and by the fact that un ≥ 1 in E, we get
Φ(1)µ(E) ≤ 1/n. This implies the claim.

370

Strongly nonlinear potential theory on metric spaces

(2) Let u ∈ B(E), then there is an open set O such that E ⊂ O and u ≥ 1 in O.
Hence

X



Φ u·Φ

−1







1
µ(E)

dµ ≥



E

Φ u·Φ

−1



1
µ(E)



dµ ≥ 1.

(5.2)

Thus
 
|u|

1
MΦ,µ
(X)

 
≥ |u|Φ,µ,X ≥

1

.
Φ−1 1/µ(E)

(5.3)

We obtain the result by taking the infimum over all u ∈ B(E).



Definition 5.2. A measure µ is said to be doubling if there is a constant C ≥ 1
such that








µ B(x,2r) ≤ Cµ B(x,r) ,

(5.4)

for every x ∈ X and r > 0.
The smallest constant C in (5.4) is called the doubling constant and is denoted
by Cd .
Theorem 5.3. Let Φ be any N-function, µ a doubling measure with the doubling
constant Cd , x0 ∈ X, and 0 < r ≤ 1. Then
 

CΦ,µ B x0 ,r



≤ 2Cd r −1

1

 
 .
Φ−1 1/µ B x0 ,r

(5.5)

Proof. We define



2r − d x,x0



,


r

u(x) = 1,















(5.6)

x ∈ B x0 ,r ,




x ∈ X B x0 ,2r .

0,

Define also



x ∈ B x0 ,2r B x0 ,r ,



1,





x ∈ B x0 ,2r ,

g(x) =  r
0,



(5.7)



x ∈ X B x0 ,2r .

We show that g ∈ D(u). Let E = B(x0 ,2r)B(x0 ,r), and x, y ∈ E. Then






d x,x0 − d y,x0
u(x) − u(y) =
r


≤

d(x, y)
.
r

(5.8)

Hence (3.12) follows in this case. Let x ∈ E and y ∈ B(x0 ,r). We have


u(x) − u(y) = 1 − u(x) =



d x,x0 − r
.
r

(5.9)

Noureddine Aı̈ssaoui 371
Since d(y,x0 ) < r ≤ d(x,x0 ), we get












d x,x0 − r ≤ d x,x0 − d y,x0 ≤ d(x, y),

(5.10)

and hence (3.12) follows. The case y ∈ E and x ∈ B(x0 ,r) is completely analogous. Now, if x, y ∈ B(x0 ,r) or x, y ∈ X B(x0 ,2r), then clearly (3.12) holds. On
the other hand, let x ∈ X B(x0 ,2r) and y ∈ B(x0 ,r). Then
u(x) − u(y) = 1 ≤

d(x, y)
,
r

(5.11)

which implies (3.12). Finally, if x ∈ E and y ∈ X B(x0 ,2r), then


u(x) − u(y) = u(x) =

2r − d x,x0
r



(5.12)

and since d(x,x0 ) < 2r < d(y,x0 ), we obtain












2r − d x,x0 ≤ d y,x0 − d x,x0 ≤ d(x, y)

(5.13)

and (3.12) holds. Thus g ∈ D(u), u ∈ B(B(x0 ,r)) and
 

CΦ,µ B x0 ,r



 
 
≤ |u|Φ,µ,X + |g |Φ,µ,X .

(5.14)

We have u ≤ χB(x0 ,2r) , |u|Φ,µ,X ≤ |χB(x0 ,2r) |Φ,µ,X . We know that


 χB(x ,2r) 
0
Φ,µ,B(x0 ,2r) =

1

 
 ,
Φ−1 1/µ B x0 ,2r

(5.15)

where χB(x0 ,2r) is the characteristic function of B(x0 ,2r).
On the other hand, g ≤ r −1 χB(x0 ,2r) , and hence
 
|g |

Φ,µ,X

≤ r −1

1

 
 .
Φ−1 1/µ B x0 ,2r

(5.16)

Thus
 

CΦ,µ B x0 ,r





≤ 1 + r −1

1

 
 .
Φ−1 1/µ B x0 ,2r

(5.17)

Now, since µ is doubling, µ(B(x0 ,2r)) ≤ Cd µ(B(x0 ,r)). Hence
1
1

 
 ≤

 
 .
Φ−1 1/µ B x0 ,2r
Φ−1 1/Cd µ B x0 ,r

(5.18)

372

Strongly nonlinear potential theory on metric spaces

It remains to evaluate Φ−1 (1/Cd µ(B(x0 ,r))). Recall that, Φ(αx) ≤ αΦ(x), for
all 0 ≤ α ≤ 1, and all x. This implies that αΦ−1 (x) ≤ Φ−1 (αx), for all 0 ≤ α ≤ 1,
and all x. Hence
1
1

 
 ≤ Cd

 
 .
Φ−1 1/Cd µ B x0 ,r
Φ−1 1/µ B x0 ,r

(5.19)

Thus
 

CΦ,µ B x0 ,r





≤ 1 + r −1 Cd

1

1/µ B x0 ,r
1
 
 .
≤ 2r −1 Cd −1 
Φ 1/µ B x0 ,r
Φ −1



 

(5.20)



This completes the proof.

5.2. Comparison between capacity and Hausdorff measures. We recall the definition of Hausdorff measures. Let h : [0,+∞[ → [0,+∞[ be an increasing function such that limr →0 h(r) = 0. For 0 < δ ≤ ∞ and E ⊂ X, we define
$

Hhδ (E) = inf

∞

 

h ri : E ⊂
i=1

∞
%




&

B xi ,ri , ri ≤ δ .

(5.21)

i=1

The h-Hausdorff measure of E is given by
Hh (E) = lim Hhδ (E).
δ →0

(5.22)

If h(t) = t s for 0 ≤ s < ∞, we obtain the s-dimensional Hausdorff measure
which we denote by Hs .
For more details on Hausdorff measures, one can consult [8, Chapter 2, Section 10].
A measure µ is regular with dimension s > 0, if there is c ≥ 1 such that, for
each x ∈ X and 0 < r ≤ diam(X),




c−1 r s ≤ µ B(x,r) ≤ cr s .

(5.23)

Note that if µ is regular with dimension s, then µ is doubling. Moreover, X has
Hausdorff dimension s and there is a constant c > 0 such that, for every E ⊂ X,
c−1 Hs (E) ≤ µ(E) ≤ cHs (E).

(5.24)

Theorem 5.4. Let Φ be an N-function satisfying the ∆2 condition and α = α(Φ) =
supt>0 (tϕ(t)/Φ(t)). Let µ be a regular measure with dimension s such that s > α.
Define h : [0,+∞[ → [0,+∞[ by h(t) = t s/α−1 . Then for every E ⊂ X,
CΦ,µ (E) ≤ cHh (E),

(5.25)

Noureddine Aı̈ssaoui 373
where the constant c depends only on Φ and the doubling constant. In particular, if
Hh (E) = 0, then CΦ,µ (E) = 0.
Proof. Let B(xi ,ri ), i = 1,2,..., be any covering of E such that the radii ri satisfy
ri ≤ 1/2, i = 1,2,.... By (5.5) we get, for all i, that
 

CΦ,µ B xi ,ri



≤ Cr −1

1

 
 .
Φ−1 1/µ B xi ,ri

(5.26)

From (5.23) we have µ(B(xi ,ri )) ≤ cris , for all i; and hence


Φ −1







1
1
1
 ≥ Φ−1
≥ c −1 Φ −1 s .
cris
ri
µ B xi ,ri
 

(5.27)

It remains to evaluate Φ−1 (1/ris ). For this goal we distinguish two cases
(i) if Φ(1) ≤ 1, then Φ−1 (1/ris ) ≥ Φ−1 (Φ(1)(1/ris ));
(ii) if Φ(1) ≥ 1, then
Φ

−1









1
1
Φ(1)
1
−1
Φ−1 Φ(1) s .
s =Φ
s ≥
ri
Φ(1)ri
Φ(1)
ri

(5.28)

Hence in the two cases we have Φ−1 (1/ris ) ≥ C  Φ−1 (Φ(1)(1/ris )). We get
Φ

−1





1
1
 ≥ C  Φ−1 Φ(1) s .
ri
µ B xi ,ri
 

(5.29)

On the other hand we know that Φ(t) ≤ Φ(1)t α , for all t ≥ 1. This implies
that t 1/α ≤ Φ−1 (Φ(1)t), for all t ≥ 1. Whence, for all i


Φ

−1



1
1
 
 ≥ C  s/α .
µ B xi ,ri
ri

(5.30)

Thus, for all i
 

CΦ,µ B xi ,ri



≤ C  r s/α−1 .

(5.31)

The subadditivity of CΦ,µ yields
CΦ,µ (E) ≤ C 

∞

 

CΦ,µ B xi ,ri
i=1



≤ C 

∞

 

h ri .

(5.32)

i=1

By taking the infimum over all coverings by balls and letting the radii tend to

zero, we finish the proof.

374

Strongly nonlinear potential theory on metric spaces

References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]

R. A. Adams, Sobolev Spaces, Academic Press, New York, 1975.
N. Aı̈ssaoui, Note sur la capacitabilité dans les espaces d’Orlicz, Ann. Sci. Math. Québec
19 (1995), no. 2, 107–113 (French).
, Bessel potentials in Orlicz spaces, Rev. Mat. Univ. Complut. Madrid 10
(1997), no. 1, 55–79.
, Instability of capacity in Orlicz spaces, Potential Anal. 6 (1997), no. 4, 327–
346.
, Some developments of strongly nonlinear potential theory, Libertas Math. 19
(1999), 155–170.
N. Aı̈ssaoui and A. Benkirane, Capacités dans les espaces d’Orlicz, Ann. Sci. Math.
Québec 18 (1994), no. 1, 1–23 (French).
, Potentiel non linéaire dans les espaces d’Orlicz, Ann. Sci. Math. Québec 18
(1994), no. 2, 105–118 (French).
H. Federrer, Geometric Measure Theory, Springer-Verlag, Berlin, 1969.
B. Franchi, P. Hajłasz, and P. Koskela, Definitions of Sobolev classes on metric spaces,
Ann. Inst. Fourier (Grenoble) 49 (1999), no. 6, 1903–1924.
D. Gallardo, Orlicz spaces for which the Hardy-Littlewood maximal operator is
bounded, Publ. Mat. 32 (1988), no. 2, 261–266.
D. Gilbarg and N. S. Trudinger, Elliptic Partial Differential Equations of Second Order,
Springer-Verlag, Berlin, 1983.
P. Hajłasz, Sobolev spaces on an arbitrary metric space, Potential Anal. 5 (1996), no. 4,
403–415.
P. Hajłasz and P. Koskela, Sobolev met Poincaré, Mem. Amer. Math. Soc. 145 (2000),
no. 688, x+101.
P. Hajłasz and O. Martio, Traces of Sobolev functions on fractal type sets and characterization of extension domains, J. Funct. Anal. 143 (1997), no. 1, 221–246.
L. I. Hedberg, On certain convolution inequalities, Proc. Amer. Math. Soc. 36 (1972),
505–510.
T. Kilpeläinen, J. Kinnunen, and O. Martio, Sobolev spaces with zero boundary values
on metric spaces, Potential Anal. 12 (2000), no. 3, 233–247.
J. Kinnunen and O. Martio, The Sobolev capacity on metric spaces, Ann. Acad. Sci.
Fenn. Math. 21 (1996), no. 2, 367–382.
M. A. Krasnosel’skiı̆ and Ja. B. Rutickiı̆, Convex Functions and Orlicz Spaces, P. Noordhoff, Groningen, 1961.
A. Kufner, O. John, and S. Fučı́k, Function Spaces, Noordhoff International Publishing, Leyden, 1977.
W. A. J. Luxemburg, Banach Function Spaces, Thesis, Technische Hogeschool te Delft,
1955.
M. M. Rao and Z. D. Ren, Theory of Orlicz Spaces, Marcel Dekker, New York, 1991.
L. Simon, Lectures on Geometric Measure Theory, Australian National University Centre for Mathematical Analysis, Canberra, 1983.
W. P. Ziemer, Weakly Differentiable Functions, Springer-Verlag, New York, 1989.

Noureddine Aı̈ssaoui: École Normale Supérieure, B.P. 5206, Ben Souda, Fès,
Morocco
E-mail address: n.aissaoui@caramail.com

Advances in

Operations Research
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Advances in

Decision Sciences
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Applied Mathematics

Algebra

Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Probability and Statistics
Volume 2014

The Scientific
World Journal
Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International Journal of

Differential Equations
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014

Submit your manuscripts at
http://www.hindawi.com
International Journal of

Advances in

Combinatorics
Hindawi Publishing Corporation
http://www.hindawi.com

Mathematical Physics
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Journal of

Complex Analysis
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International
Journal of
Mathematics and
Mathematical
Sciences

Mathematical Problems
in Engineering

Journal of

Mathematics
Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Discrete Mathematics

Journal of

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Discrete Dynamics in
Nature and Society

Journal of

Function Spaces
Hindawi Publishing Corporation
http://www.hindawi.com

Abstract and
Applied Analysis

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

International Journal of

Journal of

Stochastic Analysis

Optimization

Hindawi Publishing Corporation
http://www.hindawi.com

Hindawi Publishing Corporation
http://www.hindawi.com

Volume 2014

Volume 2014