Electronic Journal of Differential Equations, Vol. 2007(2007), No. 58, pp. 1–9.

ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu

ftp ejde.math.txstate.edu (login: ftp)

INTERVAL OSCILLATION OF SECOND-ORDER

EMDEN-FOWLER NEUTRAL DELAY DIFFERENTIAL
EQUATIONS

MU CHEN, ZHITING XU

Abstract. Employing Riccati techniques and the integral averaging method,

we establish interval oscillation criteria for the second-order Emden-Fowler
neutral delay differential equation
[|x0 (t)|γ−1 x0 (t)]0 + q1 (t)|y(t − σ)|α−1 y(t − σ) + q2 (t)|y(t − σ)|β−1 y(t − σ) = 0,
where t ≥ t0 and x(t) = y(t) + p(t)y(t − τ ). The criteria obtained here
are different from most known criteria in the sense that they are based on
information only on a sequence of subintervals of [t0 , ∞), rather than on the
whole half-line. In particular, two interesting examples that illustrate the
importance of our results are included.

1. Introduction
Consider the second-order Emden-Fowler neutral delay differential equation
[|x0 (t)|γ−1 x0 (t)]0 + q1 (t)|y(t − σ)|α−1 y(t − σ) + q2 (t)|y(t − σ)|β−1 y(t − σ) = 0, (1.1)
where t ≥ t0 and x(t) = y(t) + p(t)y(t − τ ). In what follows we assume that
(A1) τ and σ are nonnegative constants, α, β and γ are positive constants with
0 < α < γ < β;
(A2) q1 , q2 ∈ C([t0 , ∞), R+ ), R+ = (0, ∞);
(A3) p ∈ C([t0 , ∞), R), and −1 < p0 ≤ p(t) ≤ 1, p0 is a constant.
For any ϕ ∈ C([t0 −θ, t0 ], R), θ = max{τ, σ}, (1.1) has a solution y(t) extendable
on [t0 , ∞) satisfying the initial condition y(t) ≡ ϕ(t) for [t0 − θ, t0 ]; see, e.g., Hale
[6]. Our attention is restricted to those solutions y = y(t) of (1.1) which exist on
some half-line [ty , ∞) with sup{|y(t)| : t ≥ T } > 0 for any T ≥ ty , and satisfy (1.1).
As usual, a nontrivial solution of (1.1) is called oscillatory if it has arbitrarily large
zeros, otherwise it is called nonoscillatory. Finally, (1.1) is called oscillatory if all
its solutions are oscillatory. We say that (1.1) satisfies the superlinear condition if
q1 (t) ≡ 0 and it satisfies the sublinear condition if q2 (t) ≡ 0.

2000 Mathematics Subject Classification. 34K40, 34K11, 34C10.

Key words and phrases. Interval oscillation; neutral delay differential equation;
Emden-Fowler; Riccati technique; integral averaging method.
2007
c Texas State University - San Marcos.
Submitted January 5, 2007. Published April 22, 2007.
1
2 M. CHEN, Z. XU EJDE-2007/58

We note that second order neutral delay differential equation are used in many
fields such as vibrating masses attached to an elastic bar and some variational
problems, etc., see Hale [6].
In the last decades, there has been an increasing interest in obtaining sufficient
conditions for the oscillation and/or nonoscillation of second order linear and non-
linear neutral delay differential equations (see, for example, the monographs [1, 3, 5]
and the references therein). Recently, the results of Atkinson [2] and Belohorec [4]
for the Emden-Fowler equation
y 00 (t) + q(t)|y(t)|γ−1 y(t) = 0, q ∈ C([t0 , ∞), R) and γ > 0 (1.2)
have been extended to the second order neutral delay differential equation
[y(t) + p(t)y(t − τ )]00 + q(t)f ((y − σ)) = 0 (1.3)
by Wong [13] under the assumption that the nonlinear function f satisfies the
sublinear condition
Z ε Z −ε
du du
0< , < ∞ for all ε > 0,
0 + f (u) 0 − f (u)
as well as the superlinear condition
Z ∞ Z −∞
du du
0< , < ∞ for all ε > 0.
ε f (u) −ε f (u)
Also it will be of great interest to find some oscillation criteria for special case for
(1.3), even for the Emden-Fowler neutral delay differential equation
[y(t) + p(t)y(t − τ )]00 + q(t)|y(t − σ)|γ−1 y(t − σ) = 0, γ > 0. (1.4)
This problem was posed by [13, Remark d]. As an positive answer to it, Saker [10],
Saker and Manojlovic̀ [11], and Xu and Liu [14] have given some oscillation criteria
for (1.1), (1.3) and (1.4). As we know, the results obtained in [10, 11, 13, 14]
involve the integral of the functions q, q1 , q2 and hence require the information
of those functions on the the entire half-linear [t0 , ∞). As pointed out in Kong
[8], oscillation is an interval property, that is, it is more reasonable to investigate
solutions on an infinite set of bounded intervals. Therefore, the problem is to find
oscillation criteria which use only the information about the involved functions on
these intervals; outside of these intervals the behavior of the functions is irrelevant.
Such type of criteria are referred to as interval oscillation criteria. The first beautiful
interval criteria in this direction was due to Kong [8], who gave some interval criteria
for the oscillation of the second order linear ordinary differentia equation
(r(t)y 0 (t))0 (t) + q(t)y(t) = 0. (1.5)
Recently, Yang et al [15] extended Kong-type interval criteria to certain neutral
differential equations.
Motivated by the ideas of Kong [8] and Philos [9], in this paper, by using Ric-
cati technique and the integral averaging method, we will establish some interval
oscillation criteria for (1.1), that is, criteria given by the behavior of (1.1) only
on a sequence of subintervals of [t0 , ∞) (see Theorems 2.2–2.6 for details) rather
than the whole half-line. Our theorems essentially improve some known results in
[10, 14]. In particular, two interesting examples that illustrate the importance of
our results are also included.
EJDE-2007/58 INTERVAL OSCILLATION 3

2. Main results
In this section, we shall establish Kong-type interval oscillation criteria for (1.1)
under the cases when 0 ≤ p(t) ≤ 1 and −1 < p0 ≤ p(t) ≤ 0. It will be convenient
to make the following notations in the remainder of this paper. Define
β − α β − α 1
µ = min , , k=
β−γ γ−α (1 + γ)1+γ
Q1 (t) = µ[1 − p(t − σ)]γ [q1β−γ (t)q2γ−α (t)]1/(β−α) ,
Q2 (t) = µ[q1β−γ (t)q2γ−α (t)]1/(β−α) .
In the sequel, we say the a function H = H(t, s) belongs to a function class H,
denoted by H ∈ H, if H ∈ C(D, [0, ∞)), where D = {(t, s) : −∞ < s ≤ t < ∞},
and H satisfies
(H1) H(t, t) = 0, H(t, s) > 0 for t > s;
(H2) H has partial derivatives ∂H/∂t and ∂H/∂s on D such that
∂H p ∂H p
(t, s) = h1 (t, s) H(t, s) and (t, s) = −h2 (t, s) H(t, s),
∂t ∂s
where h1 , h2 ∈ Lloc (D, R).
The following Lemma will be useful for establishing oscillation criteria for (1.1)
whose proof can be found in [12].
Lemma 2.1. Let A0 , A1 , A2 ∈ C([t0 , ∞), R) with A2 > 0, and w ∈ C1 ([t0 , ∞), R).
If there exist interval (a, b) ⊂ [t0 , ∞) such that
w0 (s) ≤ −A0 (s) + A1 (s)w(s) − A2 (s)|w(s)|(γ+1)/γ , s ∈ (a, b), (2.1)
then for any c ∈ (a, b),
Z ch
1 k γγ i
H(s, a)A0 (s) − γ
|φ1 (s, a)|γ+1 ds
H(c, a) a (A2 (s))
Z bh (2.2)
1 k γγ γ+1
i
+ H(b, s)A0 (s) − |φ 2 (b, s)| ds ≤ 0
H(b, c) c (A2 (s))γ
for every H ∈ H, where
p
h1 (s, a) H(s, a) + A1 (s)H(s, a)
φ1 (s, a) = ,
(H(s, a))γ/(γ+1)
p
−h2 (b, s) H(b, s) + A1 (s)H(b, s)
φ2 (b, s) = .
(H(b, s))γ/(γ+1)
Theorem 2.2. Suppose that there exist interval (a, b) ⊂ [t0 , ∞), constant c ∈ (a, b),
and functions H ∈ H, ρ ∈ C1 ([t0 , ∞), R+ ), such that one of the following two
conditions is satisfied:
(C1) 0 ≤ p(t) ≤ 1, and
Z c
1
ρ(s) H(s, a)Q1 (s) − k|φ1 (s, a)|γ+1 ds
 
H(c, a) a
Z b (2.3)
1
ρ(s) H(b, s)Q1 (s) − k|φ2 (b, s)|γ+1 ds > 0;
 
+
H(b, c) c
4 M. CHEN, Z. XU EJDE-2007/58

(C2) −1 < p0 ≤ p(t) ≤ 0, and

Z c
1 h i
ρ(s) H(s, a)Q2 (s) − k|φ1 (s, a)|γ+1 ds
H(c, a) a
Z b (2.4)
1 h i
+ ρ(s) H(b, s)Q2 (s) − k|φ2 (b, s)|γ+1 ds > 0,
H(b, c) c
where
p
h1 (s, a) H(s, a) + (ρ0 (s)/ρ(s))H(s, a)
φ1 (s, a) = ,
(H(s, a))γ/(γ+1)
p
−h2 (b, s) H(b, s) + (ρ0 (s)/ρ(s))H(b, s)
φ2 (b, s) = .
(H(b, s))γ/(γ+1)
Then every solution of (1.1) has at least one zero in (a, b) .
Proof. Case (C1): Let y(t) be a nonoscillatory solution of (1.1). Without loss of
generality we may assume that y(t) > 0 for t ≥ t0 . (The case of y(t) < 0 can be
considered similarly). Furthermore, we may suppose that exists a t1 ≥ t0 such that

y(t) > 0, y(t − τ ) > 0, y(t − σ) > 0 for t ≥ t1 . (2.5)

As in [15, Lemma 1 (1)], for some T0 ≥ t1 + τ + σ, we have immediately that
x(t) > 0, x0 (t) > 0, x00 (t) > 0 for t ≥ T0 − τ − σ. (2.6)
Using these inequalities and noting that x(t) ≥ y(t), we obtain
y(t) = x(t) − p(t)x(t − τ ) ≥ [1 − p(t)]x(t).
Thus, for all t ≥ T0 ,
y(t − σ) ≥ [1 − p(t − σ)]x(t − σ).
Then (1.1) implies that for t ≥ T0 ,
[|x0 (t)|γ−1 x0 (t)]0 +q1 (t)[1−p(t−σ)]α xα (t−σ)+q2 (t)[1−p(t−σ)]β xβ (t−σ) ≤ 0. (2.7)
Define
(x0 (t))γ
w(t) = ρ(t) . (2.8)
xγ (t − σ)
Differentiating w(t), and using (2.7), we get
w0 (t) ≤ −ρ(t){q1 (t)[1 − p(t − σ)]α xα−γ (t − σ)
ρ0 (t)  x0 (t) γ+1
+ q2 (t)[1 − p(t − σ)]β xβ−α (t − σ)} + w(t) − γρ(t) ,
ρ(t) x(t − σ)
(2.9)
since x0 (t) < x0 (t − σ). By Young’s inequality [7, Theorem 61], we have
β−γ γ−α
q1 (t)[1 − p(t − σ)]α xα−γ (t − σ) + q2 (t)[1 − p(t − σ)]β xβ−α (t − σ)
β−α β−α
≥ [1 − p(t − σ)]γ [q1β−γ (t)q2γ−α (t)]1/(β−α) ,
and consequently,
q1 (t)[1 − p(t − σ)]α xα−γ (t − σ) + q2 (t)[1 − p(t − σ)]β xβ−α (t − σ) ≥ Q1 (t), (2.10)
Combining this inequality with (2.9), we get
ρ0 (t) γ
w0 (t) ≤ −ρ(t)Q1 (t) + w(t) − 1/γ |w(t)|(γ+1)/γ for t ≥ T0 . (2.11)
ρ(t) ρ (t)
EJDE-2007/58 INTERVAL OSCILLATION 5

Comparing the above inequality and (2.1), we find that

ρ0 (t) γ
A0 (t) = ρ(t)Q1 (t), A1 (t) = , A2 (t) = .
ρ(t) ρ1/γ (t)
Applying Lemma 2.1 to (2.11), we see that inequality (2.3) fails to hold, hence y(t)
has at least one zero in (a, b).
Case (C2): Let y(t) be a nonoscillatory solution of (1.1). Without loss of gen-
erality we may assume that y(t) > 0 for t ≥ t0 . Furthermore, as in [15, Lemma
1(2)], we suppose that there exists a t1 > t0 such that (2.5) holds. Then, for some
T0 ≥ t1 + τ + σ, we still have (2.6) holds for t ≥ T0 . Note that y(t − σ) ≥ x(t − σ)
for t ≥ T0 , (1.1) changes into
[|x0 (t)|γ−1 x0 (t)]0 + q1 (t)xα (t − σ) + q2 (t)xβ (t − σ) ≤ 0, t ≥ T0 . (2.12)
Consider the function w(t) also defined by (2.8), as in the proof of (2.11), we obtain

ρ0 (t) γ
w0 (t) ≤ −ρ(t)Q2 (t) + w(t) − 1/γ |w(t)|(γ+1)/γ . (2.13)
ρ(t) ρ (t)
The rest of proof is similar to that of case (C1) and hence is omitted. 

If the conditions of Theorem 2.2 hold for a sequence {(an , bn )} of intervals such
that limn→∞ an = ∞, then we may conclude that (1.1) is oscillatory. That is, the
following theorem is established.
Theorem 2.3. For each T ≥ t0 , if there exist H ∈ H, ρ ∈ C 1 ([t0 , ∞), R+ ), and
constants a, b, c ∈ R with T ≤ a < c < b, such that the conditions of Theorem 2.2
are satisfied, then (1.1) is oscillatory.
Theorem 2.4. Suppose that there exist H ∈ H, ρ ∈ C 1 ([t0 , ∞), R+ ), and for each
τ ≥ t0 , such that one of the following conditions is satisfied:
(C3) 0 ≤ p(t) ≤ 1, and
Z t h i
lim sup ρ(s) H(s, τ )Q1 (s) − k|φ1 (s, τ )|γ+1 ds > 0, (2.14)
t→∞ τ
Z t h i
lim sup ρ(s) H(t, s)Q1 (s) − k|φ2 (t, s)|γ+1 ds > 0; (2.15)
t→∞ τ

(C4) −1 < p0 ≤ p(t) ≤ 0, and

Z t h i
lim sup ρ(s) H(s, τ )Q2 (s) − k|φ1 (s, τ )|γ+1 ds > 0, (2.16)
t→∞ τ
Z t h i
lim sup ρ(s) H(t, s)Q2 (s) − k|φ2 (t, s)|γ+1 ds > 0, (2.17)
t→∞ τ

where φ1 and φ2 are defined in Theorem 2.2.

Then (1.1) is oscillatory.
Proof. We only prove case (C3). The proof of case (C4) is similar. For any T ≥ t0 ,
let a = T . In (2.14) we choose τ = a, then there exists c > a such that
Z c h i
ρ(s) H(s, a)Q1 (s) − k|φ1 (s, a)|γ+1 ds > 0. (2.18)
a
6 M. CHEN, Z. XU EJDE-2007/58

In (2.15) we choose τ = c. Then there exists b > c such that

Z b h i
ρ(s) H(b, s)Q1 (s) − k|φ2 (b, s)|γ+1 ds > 0. (2.19)
c

From this inequality and (2.18), we obtain (2.3). The conclusion thus comes from
Theorem 2.3 (C1). The proof is complete. 

With an appropriate choice of functions H, one can derive from Theorem 2.3 a
number of oscillation criteria for (1.1). For the case where H := H(t − s) ∈ H, we
have that h1 (t − s) ≡ h2 (t − s) and denote them by h(t − s). The subclass of H
containing such H(t − s) is denoted by H0 . Applying Theorem 2.3 to H0 , we get
the following results.
Theorem 2.5. For each T ≥ t0 , if there exist H ∈ H0 , ρ ∈ C 1 ([t0 , ∞), R+ ), and
constants a, c ∈ R with T ≤ a < c, such that one of the following conditions is
satisfied:
(C5) 0 ≤ p(t) ≤ 1, and
Z c
H(s − a)[ρ(s)Q1 (s) + ρ(2c − s)Q1 (2c − s)]ds > k Θ(a, c); (2.20)
a

(C6) −1 < p0 ≤ p(t) ≤ 0, and

Z c
H(s − a)[ρ(s)Q2 (s) + ρ(2c − s)Q1 (2c − s)]ds > k Θ(a, c), (2.21)
a

where
c
ρ0 (s)
Z γ+1
1 n  p
Θ(a, c) = ρ(s) − a) H(s − a) + H(s − a)
 
(H(s − a))γ h(s
ρ(s)

a
 p ρ0 (2c − s) γ+1 o
+ ρ(2c − s) − h(s − a) H(s − a) + H(s − a) ds.
 
ρ(2c − s)
Then (1.1) is oscillatory.
Let
H(t, s) = (t − s)λ , (t, s) ∈ D, (2.22)
where λ > max{1, γ} is a constant. Then H ∈ H0 and
λ−2
h1 (t − s) = h2 (t − s) = h(t − s) = λ(t − s) 2 .
By Theorem 2.4, we obtain the following oscillation criteria of Kamenev’s type.
Theorem 2.6. For each τ ≥ t0 and for some λ > max{1, γ}. Then (1.1) is
oscillatory provided that one of the following conditions is satisfied:
(C7) 0 ≤ p(t) ≤ 1, and
t
k λγ+1
Z
1
lim sup (s − τ )λ Q1 (s)ds > , (2.23)
t→∞ tλ−γ τ λ−γ
Z t
1 k λγ+1
lim sup λ−γ (t − s)λ Q1 (s)ds > ; (2.24)
t→∞ t τ λ−γ
EJDE-2007/58 INTERVAL OSCILLATION 7

(C8) −1 < p0 ≤ p(t) ≤ 0, and

t
k λγ+1
Z
1
lim sup (s − τ )λ Q2 (s)ds > , (2.25)
t→∞ tλ−γ τ λ−γ
Z t
1 k λγ+1
lim sup λ−γ (t − s)λ Q2 (s)ds > . (2.26)
t→∞ t τ λ−γ
Proof. We only prove Case (C7). The proof of Case (C8) is similar. Taking H(t, s)
as in (2.22) and ρ(t) ≡ 1 for t ≥ t0 , we get
Z t Z t
λγ+1
ρ(s)|φ1 (s, τ )|γ+1 ds = λγ+1 (s − τ )λ−(γ+1) ds = (t − τ )λ−γ ,
τ τ λ−γ
Z t Z t
λγ+1
ρ(s)|φ2 (t, s)|γ+1 ds = λγ+1 (t − s)λ−(γ+1) ds = (t − τ )λ−γ .
τ τ λ−γ
In view of the fact λ > γ, it follows that
Z t
1 λγ+1
lim λ−γ ρ(s)|φ1 (s, τ )|γ+1 ds = , (2.27)
t→∞ t τ λ−γ
Z t
1 λγ+1
lim ρ(s)|φ2 (t, s)|γ+1 ds = , (2.28)
t→∞ tλ−γ τ λ−γ
From (2.23) and (2.27), we have
Z t
1
lim sup λ−γ ρ(s)[H(s, τ )Q1 (s) − k|φ1 (s, τ )|γ+1 ]ds
t→∞ t τ
Z t
1 k λγ+1
= lim sup λ−γ (s − τ )λ Q1 (s)ds − > 0.
t→∞ t τ λ−γ
i.e., (2.14) holds. Similarly, (2.24) and (2.28) imply that (2.15) holds, By Theorem
2.4 (C3), (1.1) is oscillatory. The proof is complete. 

3. Examples
In final section, we will show the application of our oscillation criteria by two
examples.
Example 3.1. Consider the equation
 0 γ−1 0 0
|x (t)| x (t) + q1 (t)|y(t − 2)|α−1 y(t − 2) + q2 (t)|y(t − 2)|β−1 y(t − 2) = 0, (3.1)
where t ≥ 2 and x(t) = y(t) + p0 y(t − 1), −1 < p0 < 1, 0 < α < γ < β with
γ = (α + β)/2 and q1 , q2 ∈ C([t0 , ∞), R+ ) with

η(t − 3n),
 3n < t ≤ 3n + 1,
q1 (t) = q2 (t) = η(−t + 3n + 2), 3n + 1 < t < 3n + 2,

q0 (t), 3n + 2 ≤ t ≤ 3n + 3,


for n ∈ {1, 2, . . . }, where q0 (t) is any positive continuous function which makes
q1 (t) a continuous function, and
k(λ + 2)λγ+1

 2(1 − p )γ (λ − γ) , 0 ≤ p0 < 1,


0
η>
 k(λ + 2)λγ+1

 , −1 < p0 ≤ 0,
2(λ − γ)
8 M. CHEN, Z. XU EJDE-2007/58

is a constant for fixed λ > max {1, γ}. Now, we consider the following two cases:
Case 1: 0 ≤ p0 < 1. Note that Q1 (t) = 2(1 − p0 )γ q1 (t). Let a = 3n, c = 3n + 1,
H(t, s) = (t − s)λ and ρ(t) ≡ 1. Then
Z c Z c
(s − a)λ [Q1 (s) + Q1 (2c − s)]ds − 2kλγ+1 (s − a)λ−γ−1 ds
a a
Z 3k+1 Z 3k+1
≥ 4η(1 − p0 )γ (s − 3k)γ+1 ds − 2kλγ+1 (s − 3k)λ−(γ+1) ds
3k 3k
4η(1 − p0 )γ 2kλγ+1
= − > 0,
λ+2 λ−γ
i.e., (2.20) holds.
Case 2: −1 < p0 ≤ 0. Note that Q2 (t) ≥ 2q1 (t). The rest of proof is similar to
that of Case 1. Thus, (3.1) is oscillatory by Theorem 2.5.
However, the known results such as in [10, 11, 13, 14, 15] do not apply to (3.1).

Example 3.2. Consider the neutral delay differential equation

[|x0 (t)|γ−1 x0 (t)]0 + q1 (t)|y(t − 2)|α−1 y(t − 2) + q2 (t)|y(t − 2)|β−1 y(t − 2) = 0, (3.2)
where t ≥ 2 and x(t) = y(t) + p0 y(t − 1), −1 < p0 < 1, 0 < α < γ < β with
γ = (α + β)/2, and q1 , q2 ∈ C([t0 , ∞), R+ ) with q1 (t)q2 (t) ≥ (ε/tγ+1 )2 with

kγ0γ+1

, 0 ≤ p0 < 1,


2(1 − p0 )γ

ε>
γ+1
 kγ0 ,

−1 < p0 ≤ 0,

2
where γ0 = max{1, γ}. Let ρ(t) = t−(γ+1) . The following two cases will be consid-
ered.
Case 1: 0 ≤ p0 < 1. Note that Q1 (t) ≥ 2ε (1 − p0 )γ t−(γ+1) . Let λ > γ0 . Then
Z t Z t
1 λ γ 1 (s − τ )λ
lim λ−γ (s − τ ) Q1 (s)ds ≥ 2ε(1 − p0 ) lim λ−γ ds
t→∞ t τ t→∞ t τ sγ+1
2ε(1 − p0 )γ (t − τ )λ
= lim
λ−γ t→∞ tλ
γ
2ε(1 − p0 )
= .
λ−γ

For any ε > k γ0γ+1 /(2(1 − p0 )γ ), there exists λ > γ0 such that
2ε(1 − p0 )γ k λγ+1
> .
λ−α λ−α
This means that (2.23) holds. By [8, Lemma 3.1], (2.24) holds for the same λ.
Applying Theorem 2.6 (C7), we find that (3.2) is oscillatory.
Case 2: −1 < p0 ≤ 0. Note that Q2 (t) ≥ 2ε t−(γ+1) . The rest of proof is similar
to that of Case 1. Hence, Theorem 2.6 (C8) holds. Thus, (3.2) is oscillatory .

Acknowledgments. The authors would like to express his great appreciation to

Professor Qingkai Kong for many valuable suggestions and useful comments.
EJDE-2007/58 INTERVAL OSCILLATION 9

References
[1] R. P. Agarwal, S. R. Grace, D. O’Regan; Oscillation Theory for Second Order Dynamic
Equations, Taylor & Francis, London, 2003.
[2] F. V. Atkinson; On second order nonlinear oscillation, Pacific J. Math. 5 (1955) 643-647.
[3] D. D. Bainov, D. P. Mishev; Oscillation Theory for Neutral Equations with delay, Adam
Hilger IOP Publishing Ltd., 1991.
[4] S. Belohorec; Oscillatory solution of certain nonlinear differential equations of the second
order, Math. Fyz. Casopis Sloven. Akad. Vied. 11 (1961) 250-255.
[5] L. H. Erbe, Q. Kong, B. G. Zhang; Oscillation Theory for Functional Differential Equations,
Marcel Dekker, New York, 1995.
[6] J. K. Hale; Theory of Functional Differential Equations, Springer-Verlag, New York, 1977.
[7] G. H. Hardy, J. E. Littlewood, G. Polya; Inequalities, second ed., Cambridge University Press,
Cambridge, 1988.
[8] Q. Kong; Interval criteria for oscillation of second order linear ordinary differential equa-
tions, J. Math. Anal. Appl. 229 (1999) 258-270.
[9] Ch. G. Philos; Oscillation theorems for linear differential equations of second order, Arch.
Math. (Besel). 53 (1989) 482-492.
[10] S. H. Saker; Oscillation for second order neutral delay differential equations of Emden-Fowler
type, Acta. Math. Hungar. 100 (1-2) (2003) 37-62.
[11] S. H. Saker, J. V. Manojlivic̀; Oscillation criteria for second order superlinear neutral delay
differential equations, Electron. J. Qual. Theory Differ. Equ. 10 (2004) 1-22.
[12] A. Tiryaki, Y. Basci, I. Gülec; Interval criteria for oscillation of second order functional
differential equations, Comput. Math. Appl. 59 (2005) 1487-1498.
[13] J. S. W. Wong; Necessary and sufficient conditions for oscillation for second order neutral
differential equations, J. Math. Anal. Appl. 252 (2000) 342-352.
[14] Z. Xu, X. Liu; Philos-type oscillation criteria for Emden-Fowler neutral delay differential
equations, J. Comput. Appl. Math. doi:10.1016/j.cam.2006. 09.012.
[15] Q. Yang, L. Yang, S. Zhu; Interval criteria for oscillation of second order nonlinear neutral
differential equations, Comput. Math. Appl. 46 (2003) 903-918.

Mu Chen
School of Mathematical Sciences, South China Normal University, Guangzhou, 510631,
China.
Department of Mathematics, Zhanjiang Education College, Zhanjiang, Guangdong 524037,
China
E-mail address: zjchenmu03@163.com

Zhiting Xu
School of Mathematical Sciences, South China Normal University, Guangzhou, 510631,
China
E-mail address: xztxhyyj@pub.guangzhou.gd.cn