Electronic Journal of Qualitative Theory of Differential Equations

2007, No. 21, 1-10; http://www.math.u-szeged.hu/ejqtde/

On global attractivity of solutions of a functional-integral equation

Mohamed Abdalla Darwish

Department of Mathematics, Faculty of Science,

Alexandria University at Damanhour, 22511 Damanhour, Egypt
e-mail: [email protected], [email protected]

Abstract

We prove an existence theorem for a quadratic functional-integral equation

of mixed type. The functional-integral equation studied below contains as spe-
cial cases numerous integral equations encountered in nonlinear analysis. With
help of a suitable measure of noncompactness, we show that the functional inte-
gral equation of mixed type has solutions being continuous and bounded on the
interval [0, ∞) and those solutions are globally attractive.

Key words and phrases: Functional-integral equation, existence, global attractivity,

measure of noncompactness, fixed point theorem due to Darbo.
AMS (MOS) Subject Classifications: 45G10, 45M99, 47H09.

1 Introduction

Quadratic integral equations are often applicable in the theory of radiative transfer,
kinetic theory of gases, in the theory of neutron transport and in the traffic theory.
Especially, the so-called quadratic integral equation of Chandrasekhar type can be very
often encountered in many applications (cf. [1, 2, 3, 6, 7, 8, 9, 10, 13, 14]). In this
paper we study the functional integral equation of mixed type, namely
 Z t Z ∞ 
x(t) = f t, x(t), u(t, s, x(s))ds, v(t, s, x(s)) ds . (1)
0 0

Eq.(1) contains as special cases numerous integral and functional-integral equations

encountered in nonlinear analysis. Also, the famous Chandrasekhar’s integral equation
is considered as a special case.
Using the technique associated with a suitable measure of noncompactness, we
show that Eq.(1) has solutions being continuous and bounded on the interval [0, ∞)
and those solutions are globally attractive. In fact, our result in this paper is motivated
by the extension of the work of Hu and Yan [12].

EJQTDE, 2007 No. 21, p. 1

2 Auxiliary facts and results

This section collects some definitions and results which will be needed further on.
Assume that (E, k.k) is a Banach space with zero element θ. Let B(x, r) denote the
closed ball centered at x and with radius r. The symbol Br stands for the ball B(θ, r).
If X is a subset of E, then X̄ and ConvX denote the closure and convex closure of
X, respectively. Moreover, we denote by ME the family of all nonempty and bounded
subsets of E and NE its subfamily consisting of all relatively compact subsets.
Next we give the concept of a measure of noncompactness [5]:
Definition 2.1 A mapping µ : ME → [0, +∞) is said to be a measure of noncom-
pactness in E if it satisfies the following conditions:
1) The family Kerµ = {X ∈ ME : µ(X) = 0} is nonempty and Kerµ ⊂ NE .
2) X ⊂ Y ⇒ µ(X) ≤ µ(Y ).
3) µ(X̄) = µ(ConvX) = µ(X).
4) µ(λX + (1 − λ)Y ) ≤ λ µ(X) + (1 − λ) µ(Y ) for 0 ≤ λ ≤ 1.
5) If Xn ∈ ME , Xn = X̄n , Xn+1 ⊂ Xn for n = 1, 2, 3, ... and lim µ(Xn ) = 0 then
n→∞
∩∞n=1 Xn 6= φ.

We recall the fixed point theorem due to Darbo [11]. Before quoting this theorem,
we need the following definition:
Definition 2.2 Let M be a nonempty subset of a Banach space E, and T : M → E
be a continuous operator that transforms bounded sets onto bounded ones. We say that
T satisfies the Darbo condition (with constant c ≥ 0) with respect to a measure of
noncompactness µ if for any bounded subset X of M we have
µ(TX) ≤ c µ(X).
If T satisfies the Darbo condition with c < 1 then it is called a contraction operator
with respect to µ.
Theorem 2.1 Let Q be a nonempty, bounded, closed and convex subset of the space
E and let
H:Q→Q
be a contraction with respect to the measure of noncompactness µ.
Then H has a fixed point in the set Q.
Remark 2.1 [5]
Under the assumptions of the above theorem the set Fix H of fixed points of H be-
longing to Q is a member of ker µ. In fact, as µ(H(Fix H)) = µ(Fix H) ≤ c µ(Fix H)
and 0 ≤ c < 1, we deduce that µ(Fix H) = 0.

EJQTDE, 2007 No. 21, p. 2

This observation allows us to characterize solutions of considered operator equation.
In what follows we will work in the Banach space BC(IR+ ) consisting of all real
functions defined, bounded and continuous on IR+ . The space BC(IR+ ) is equipped
with the standard norm
k x k = sup{|x(t)| : t ≥ 0}
Now, we recollect the construction of the measure of noncompactness which will be
used in the next section (see [4])
Let us fix a nonempty and bounded subset X of BC(IR+ ) and let T be a positive
number. For x ∈ X and ε ≥ 0 denote by ω T (x, ε) the modulus of continuity of the
function x on the interval [0, T ], i.e.,

ω T (x, ε) = sup{|x(t) − x(s)| : t, s ∈ [0, T ], |t − s| ≤ ε}

Further, let us put

ω T (X, ε) = sup{ω T (x, ε) : x ∈ X},

ω0T (X) = lim ω T (X, ε),
ε→0
ω0 (X) = lim ω0T (X).
T →∞

For a fixed number t ≥ 0 we denote

X(t) = {x(t) : x ∈ X}

and
diamX(t) = sup{|x(t) − y(t)| : x, y ∈ X}.
Now, let us define the function µ on the family MBC(IR+ ) by the following formula

µ(X) = ω0 (X) + lim sup diamX(t).

t→∞

It can be shown [4] that the function µ is a measure of noncompactness on the space
BC(IR+ ).

Definition 2.3 The solution x(t) of Eq.(1) is said to be globally attractive, if there
are
lim (x(t) − y(t)) = 0
t→+∞

for any solution y(t) of Eq.(1).

3 Main Results
In this section, we will study Eq.(1) assuming that the following assumptions are
satisfied:

EJQTDE, 2007 No. 21, p. 3

 a1 ) f : IR+ × IR+ × IR × IR → IR is continuous and the function t → f (t, 0, 0, 0) is
an element of the space BC(IR+ ).

a2 ) There exist continuous functions m1 , m2 : IR+ → IR+ and a constant 0 ≤ k < 1

such that

|f (t, x1 , y1 , z1 ) − f (t, x2 , y2, z2 )| ≤ k |x1 − x2 | + m1 (t)|y1 − y2 | + m2 (t)|z1 − z2 |

for all xi , yi , zi ∈ IR; i = 1, 2 and t ∈ IR+

a3 ) u, v : IR+ × IR+ × IR → IR are continuous functions such that

Z t
lim m1 (t) |u(t, s, x(s))| ds = 0,
t→∞ 0
Z ∞
lim m2 (t) |v(t, s, x(s))| ds = 0
t→∞ 0

uniformly with respect to x ∈ BC(IR+ ).

Now, we are in a position to state and prove our main result in the paper

Theorem 3.1 Let the assumptions a1 ) − a3 ) be satisfied. Then Eq.(1) has at least
one solution x ∈ BC(IR+ ) which is globally attractive.

Proof: Denote by F the operator associated with the right-hand side of Eq.(1), i.e.,
equation (1) takes the form
x = F x, (2)
where
 Z t Z ∞ 
(F x)(t) = f t, x(t), u(t, s, x(s)) ds, v(t, s, x(s)) ds , t ∈ IR+ (3)
0 0

Solving Eq.(1) is equivalent to finding a fixed point of the operator F defined on the
space BC(IR+ ).
Clearly, in virtue of our assumptions the function F x is continuous on the interval
IR+ for any function x ∈ BC(IR+ ). Moreover, from our assumptions we have
  Z t Z ∞  
 
|(F x)(t)| ≤ f t, x(t),
 u(t, s, x(s)) ds, v(t, s, x(s)) ds − f (t, 0, 0, 0))
0 0
+ |f (t, 0, 0, 0)|
Z t Z ∞
≤ k |x(t)| + m1 (t) |u(t, s, x(s))| ds + m2 (t) |v(t, s, x(s))| ds
0 0
+ |f (t, 0, 0, 0)|

Hence,
kF xk ≤ k kxk + A, (4)

EJQTDE, 2007 No. 21, p. 4

where
 Z t
A = sup m1 (t) |u(t, s, x(s))| ds+
0
Z ∞ 
m2 (t) |v(t, s, x(s))| ds + |f (t, 0, 0, 0)| : t ∈ IR+ ∈ IR+
0

A
Since k < 1, this implies that F (Br ) ⊂ Br for r = 1−k .
We claim that the operator F is continuous on Br . To establish this claim, let us
fix ε > 0 and take arbitrary x, y ∈ Br such that kx − yk ≤ ε. Then, for t ∈ IR+ , we
have
t ∞
  Z Z 

|(F x)(t) − (F y)(t)| ≤ f t, x(t), u(t, s, x(s)) ds, v(t, s, x(s)) ds
0 0
 Z t Z ∞ 

−f t, y(t), u(t, s, y(s)) ds, v(t, s, y(s)) ds 
0 0
Z t
≤ k |x(t) − y(t)| + m1 (t) |u(t, s, x(s)) − u(t, s, y(s))| ds
0
Z ∞
+m2 (t) |v(t, s, x(s)) − v(t, s, y(s))| ds (5)
0

By assumptions a1)−a3) we choose T > 0 such that for t ≥ T the following inequalities
hold Z t
ε
m1 (t) |u(t, s, x(s))| ds ≤ (1 − k) (6)
0 4
and
ε
Z ∞
m2 (t) |v(t, s, x(s))| ds ≤ (1 − k). (7)
0 4
Thus for t ≥ T , in view of (5)-(7) we obtain
ε ε ε ε
|(F x)(t) − (F y)(t)| ≤ k ε + (1 − k) + (1 − k) + (1 − k) + (1 − k)
4 4 4 4
= ε

On the other hand, taking into account the uniform continuity of the functions
u = u(t, s, x) and v = v(t, s, x) on the set [0, T ] × [0, T ] × [−r, r], we deduce that
ω1 (ε), ω2 (ε) → 0 as ε → 0, where

ω1 (ε) = sup {|u(t, s, x) − u(t, s, y)| : t, s ∈ [0, T], x, y ∈ [−r, r], |x − y| ≤ ε}

and

ω2 (ε) = sup {|u(v, s, x) − v(t, s, y)| : t, s ∈ [0, T], x, y ∈ [−r, r], |x − y| ≤ ε} .

Thus,

|(F x)(t) − (F y)(t)| ≤ k ε + T ω1 (ε) sup {m1 (t) : t ∈ [0, T ]}

+ω2 (ε) sup {m2 (t) : t ∈ [0, T ]}

EJQTDE, 2007 No. 21, p. 5

Hence, we established our claim.
Now, let X be a nonempty set of Br . Then, for x, y ∈ X and t ∈ IR+ , we get
Z t
|(F x)(t) − (F y)(t)| ≤ k |x(t) − y(t)| + m1 (t) |u(t, s, x(s)) − u(t, s, y(s))| ds
0
Z ∞
+m2 (t) |v(t, s, x(s)) − v(t, s, y(s))| ds
0
Z t
≤ k |x(t) − y(t)| + m1 (t) |u(t, s, x(s))| ds
0
Z t Z ∞
+m1 (t) |u(t, s, y(s))| ds + m2 (t) |v(t, s, x(s))| ds
Z0 ∞ 0

+m2 (t) |v(t, s, y(s))| ds

0

Hence

diam(F X)(t) ≤ k diamX(t)

 Z t Z t
+ sup m1 (t) |u(t, s, x(s))| ds + m1 (t) |u(t, s, y(s))| ds
x, y∈X 0 0
Z ∞ Z ∞ 
+m2 (t) |v(t, s, x(s))| ds + m2 (t) |v(t, s, y(s))| ds
0 0

Therefore,
lim sup diam(F X)(t) ≤ k lim sup diamX(t), (8)
t→∞ t→∞

thanks to assumption a3 ).
For any T > 0 and ε > 0. Choose a function x ∈ X and t, s ∈ [0, T] such that
|t − s| ≤ ε, without loss of generality we may assume that t > s. Then, in view of our
assumptions we have

|(F x)(t) − (F x)(s)| ≤ k |x(t) − x(s)|

Z
 t Z s 

+m1 (t)  u(t, τ, x(τ )) dτ − u(s, τ, x(τ )) dτ 
Z0∞ 0
Z ∞ 
 
+m2 (t)  v(t, τ, x(τ )) dτ − v(s, τ, x(τ )) ds
0 0
  Z s Z ∞ 

+ f t, x(s), u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ
0 0
 Z s Z ∞ 

−f s, x(s), u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ 
0 0
≤ k |x(t) − x(s)|
t
Z 
 
+m1 (t)  u(t, τ, x(τ )) dτ 
Zs s Z s 
 
+m1 (t)  u(t, τ, x(τ )) dτ − u(s, τ, x(τ )) dτ 
Z0∞ 0
Z ∞ 
 
+m2 (t)  v(t, τ, x(τ )) dτ − v(s, τ, x(τ )) ds
0 0

EJQTDE, 2007 No. 21, p. 6

   Z s Z ∞ 

+ f t, x(s),
 u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ
0 0
 Z s Z ∞ 

−f s, x(s), u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ 
0 0

Hence
ω T (Fx, ε) ≤ k ω T (x, ε)
+m1 (t) ε sup {|u(t, s, x(s))| : t, s ∈ [0, T ], kxk ≤ r}
+m1 (t) T sup {|u(t, τ, x(τ )) − u(s, τ, x(τ ))| : t, s, τ ∈ [0, T ], kxk ≤ r}
+m2 (t) sup {|v(t, τ, x(τ )) − v(s, τ, x(τ ))| : t, s, τ ∈ [0, T ], kxk ≤ r}
 

Z s Z ∞ 
+ sup f t, x(s), u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ
0 0
 Z s Z ∞ 


−f s, x(s), u(s, τ, x(τ )) dτ, v(s, τ, x(τ )) dτ  : t, s ∈ [0, T ], kxk ≤ r
0 0

Since f (t, x, y, z) is uniformly continuous on the set [0, T ]×[−r, r]×[−N, N]×[−M, M]
and the functions u(t, s, x) and v(t, s, x) are uniformly continuous on the set [0, T ] ×
[0, T ] × [−r, r], where
Z s 
N = sup u(s, τ, x(τ )) dτ : s ∈ [0, T ], kxk ≤ r
0

and Z ∞ 
M = sup v(s, τ, x(τ )) dτ : s ∈ [0, T ], kxk ≤ r ,
0
we have
sup {|f (t, x, y, z) − f (s, x, y, z)| : t, s ∈ [0, T ], |s − t| ≤ ε, kxk ≤ r, ; kyk ≤ N, kzk ≤ M }
→ 0 as ε → 0,

sup {|u(t, τ, x) − u(s, τ, x)| : t, s, τ ∈ [0, T ], |s − t| ≤ ε, kxk ≤ r} → 0 as ε → 0

and

sup {|v(t, τ, x) − v(s, τ, x)| : t, s, τ ∈ [0, T ], |s − t| ≤ ε, kxk ≤ r} → 0 as ε → 0.

Hence
ω0 (F X) ≤ k ω0 (X). (9)
From (8) and (9), keeping in mind the definition of the measure of noncompactness in
the space BC(IR+ ), we obtain

µ(F X) ≤ k µ(X). (10)

The above obtained inequality together with the fact that k < 1 enable us to apply
Theorem 2.1. Hence, we infer that Eq.(1) has at least one solution x(t).

EJQTDE, 2007 No. 21, p. 7

 Now, for any other solution y(t) of Eq.(1), we have
|x(t) − y(t)| = |(Fx)(t) − (Fy)(t)|
Z t
≤ k |x(t) − y(t)| + m1 (t) |u(t, s, x(s))| ds
0
Z t Z ∞
+m2 (t) |u(t, s, y(s))| ds + m2 (t) |v(t, s, x(s))| ds
Z0 ∞ 0

+m2 (t) |v(t, s, y(s))| ds.

0

Thus
1
 Z t Z t 
lim |x(t) − y(t)| ≤ lim m1 (t) |u(t, s, x(s))| ds + |u(t, s, y(s))| ds
t→∞ 1−k t→∞ 0 0
Z ∞ Z t 
+ lim m2 (t) |v(t, s, x(s))| ds + |v(t, s, y(s))| ds
t→∞ 0 0
= 0.

This complete the proof.

4 Examples
Example 4.1 If f (t, x, y, z) = g(t)+y, then Eq.(1) is the well-known Urysohn-Volterra
integral equation Z t
x(t) = g(t) + u(t, s, x(s)) ds.
0
On the other hand, for f (t, x, y, z) = h(t)+z, Eq.(1) reduces to the well-known Urysohn
integral equation Z ∞
x(t) = h(t) + v(t, s, x(s)) ds.
0

Example 4.2 If f (t, x, y, z) = g(t, x, y), then Eq.(1) becomes a functional-integral

equation  Z t 
x(t) = g t, x(t), u(t, s, x(s)) ds . (11)
0
In [12], the authors proved the existence of solutions to Eq.(11). These solutions
continuous and bounded on the interval [0, ∞) and are globally attractive.
t
Example 4.3 In the case f (t, x, y, z) = 1 + x z and v(t, s, x) = t+s φ(s) x, Eq.(1) has
the form
t
Z ∞
x(t) = 1 + x(t) φ(s) x(s) ds. (12)
0 t+s
Eq.(12) creates an unbounded version of the famous quadratic integral equation of
Chandrasekhar type
Z 1 t
x(t) = 1 + x(t) φ(s) x(s) ds. (13)
0 t+s

EJQTDE, 2007 No. 21, p. 8

Eq.(13) considered in many papers and monographs (cf. [1, 3, 8, 13] for instance).
Some Problems considered in the theory of radiative transfer, in the theory of neutron
transport and in the kinetic theory of gases lead to Eq.(13) (cf. [2, 3, 6, 7, 8, 9, 10, 13,
14]).

Remark 4.1 In order to apply our technique to Eq.(12) we have to impose an addi-
tional condition that the characteristic function φ is continuous and satisfies φ(0) = 0.
This condition will ensure that the kernel v(t, s, x) defined by
(
0, s = 0, t ≥ 0, x ∈ IR
v(t, s, x) = t
t+s
φ(s) x, s 6= 0, t ≥ 0, x ∈ IR

is continuous on I × I × IR in accordance with assumption a3 ), see [6].

Acknowledgement. The author would like to thank the referee for his corrections and
valuable remarks. This work was completed when the author was visiting the ICTP,
Trieste, Italy as a Regular Associate. It is a pleasure for him to express gratitude for
its financial support and the warm hospitality.

References
[1] I.K. Argyros, Quadratic equations and applications to Chandrasekhar’s and re-
lated equations, Bull. Austral. Math. Soc. 32 (1985), 275-292.

[2] J. Banaś and B. Rzepka, On existence and asymptotic stability of a nonlinear

integral equation, J. Math. Anal. Appl. 284 (2003), 165-173.

[3] J. Banaś, M. Lecko and W.G. El-Sayed, Existence theorems of some

quadratic integral equations, J. Math. Anal. Appl. 222 (1998), 276-285.

[4] J. Banaś, Measures of noncompactness in the space of continuous tempered func-

tions, Demonstratio Math. 14 (1981), 127-133.

[5] J. Banaś and K. Goebel, Measures of noncompactness in Banach spaces, Lec-

ture Notes in Pure and Applied Mathematics 60, Marcel Dekker, New York, 1980.

[6] J. Caballero, A.B. Mingarelli and K. Sadarangani, Existence of so-

lutions of an integral equation of Chandrasekhar type in the theory of radiative
transfer, Electronic Journal of Differential Equations 2006 (57) (2006), 1-11.

[7] K.M. Case and P.F. Zweifel, Linear Transport Theory, Addison-Wesley,
Reading, MA 1967.

[8] S. Chandrasekher, Radiative Transfer, Dover Publications, New York, 1960.

EJQTDE, 2007 No. 21, p. 9

 [9] M.A. Darwish, On quadratic integral equation of fractional orders, J. Math.
Anal. Appl. 311 (2005) 112-119.

[10] K. Deimling, Nonlinear fuctional analysis, Springer-Verlag, Berlin, 1985.

[11] J. Dugundji and A. Granas, Fixed Point Theory, Monografie Mathematyczne,

PWN, Warsaw, 1982.

[12] X. Hu, J. Yan, The global attractivity and asympototic stability of solution of a
nonlinear integral equation, J. Math. Anal. Appl. 321 (2006) 147-156.

[13] S. Hu, M. Khavani and W. Zhuang, Integral equations arrising in the kinetic
theory of gases, Appl. Analysis, 34 (1989), 261-266.

[14] C.T. Kelly, Approximation of solutions of some quadratic integral equations in

transport theory, J. Integral Eq. 4 (1982), 221-237.

(Received July 23, 2007)

EJQTDE, 2007 No. 21, p. 10