On the Oscillation of Second-Order Nonlinear Neutral Delay Dynamic Equations on Time Scales ()
Received 9 May 2016; accepted 13 June 2016; published 16 June 2016

1. Introduction
The theory of time scales was first proposed by Hilger [1] in order to unify continuous and discrete analysis. Several researchers have made greater contributions to various aspects of this new theory; see [2] - [4] . The new theory of dynamic equations on time scales not only unifies the theories of differential equations and difference equations, but also extends these classical cases to cases “in between”, e.g., to so-called q-difference equations where
.
In recent years, there has been much research involving the oscillation and nonoscillation of solutions of various equations on time scales such as [5] - [18] . In this paper we study and give the sufficient conditions for oscillation of the second-order neutral delay dynamic equation
(1.1)
where
and
is unbounded time scale. Besides that, we will have hypotheses as follows throughout the paper:
(H1)
is the ratio of two positive odd integers and
.
(H2)
are positive rd-continuous functions with
satisfying
.
(H3)
is a strictly increasing and differentiable function such that
, 
as
and
.
(H4)
is a continuous function which satisfies
for all
where L is a positive constant.
In addition, for the sake of clearness and convenience,we will use the notation
![]()
in the following narrative.
It is well known by reserchers in this field that an dynamic equation is called oscillatory in case all its solutions are oscillatory, and a solution of the equation is said to be oscillatory if it is neither eventually positive nor eventually negative. We only discuss those solutions x of Equation (1.1) that are not eventually zero in this paper. Moreover we refer to [3] [4] for general basic background, ideas and more details on dynamic equations.
Because of
, we shall consider Equation (1.1) respectively based on the case
(1.2)
and the other case
(1.3)
2. Several Lemmas
In this section, we present and prove three lemmas which play important roles in the proofs of the main results.
Lemma 1. ( [16] ) Assume that
is strictly increasing,
is a time scale and
. Let
. If
and
exist for
, then
exists, and
(2.1)
Lemma 2. ( [3] ) Assume that x is delta-differentiable and eventually positive or eventually negative, then
(2.2)
We give the below lemma and prove it similar to that of Q. Zhang and X. Song ( [17] , Lemma 3.5).
Lemma 3. Based on (1.2), assume that (H1)-(H4) hold. If x is an eventually positive solution of (1.1), there exists
such that
(2.3)
Proof. Assume
is an eventually positive solution of (1.1). That is, there exists
such that
and
for
Because of
and
, we get
esaily for
. At the same time for
, from equation (1.1) we obtain that
(2.4)
so
is decreasing. From (2.4), we know that
is either eventually positive or eventually negative. Now we assert that
.
Suppose to the contrary that there exits
such that
for all
. Because
is decreasing,
(2.5)
for
, where
. Based on the above inequality (2.5), we get
(2.6)
After integrating the two sides of inequality (2.6) from t2 to
, we have
(2.7)
When
, we get
from (1.2) and the above (2.7), which is contradictory to
. So the above hypothesis of
is false. In other words, we get
for
. This completes the proof. □
3. Main Results
Now we state and prove our main results in this section.
Theorem 1. Based on (1.2), assume that the conditions (H1)-(H4) hold. If there exists a positive nondecreasing D-differentiable function
such that for every ![]()
(3.1)
where
(3.2)
then (1.1) is oscillatory on
.
Proof. Assume that (1.1) has a nonoscillatory solution x on
. We may assume that
and
for all
. By the definition of
, it follows
. From (H3) we
know
, by Lemma 3 we have
, so
, and
, we obtain
![]()
The proof that x is eventually negative is similar. By Lemma 3 we have
for all
,
, and by Lemma 1 and (H3), there exists
such that
for all
.
Using (2.2) and (2.3), we have
![]()
is unbounded above, which implies
. Furthermore, from Lemma 1 we get
![]()
Thus, by (H3),
(3.3)
Next we define the function
by
(3.4)
Then on
, we have
. From the basic knowledge of the time scale calculus that you can see in [3] , we obtain
![]()
From (1.1) and (H4), we get
![]()
i.e.,
(3.5)
![]()
![]()
(3.6)
On the other hand, because
![]()
we get
(3.7)
Using (3.7) in (3.6), we have
(3.8)
At last, integrating (3.8) from T to t, we obtain
![]()
which creates a contradiction to (3.1). This completes the proof. □
Remark 1. From Theorem 1, we can obtain different conditions for oscillation of all solutions of (1.1) with different choices of
.
Next, we give the conditions that guarantee every solution of (1.1) oscillates when (1.3) holds.
Theorem 2. Based on (1.3), assume that the conditions (H1)-(H4), (3.1) and (3.2) hold. If for every ![]()
(3.9)
where
(3.10)
then (1.1) is oscillatory on
.
Proof. Assume that (1.1) has a nonoscillatory solution x on
, then it is neither eventually positive nor eventually negative. Without loss of generality, we may assume that
, then
for all
,
, it follows
and
![]()
The proof is similar when x is eventually negative. Since
is decreasing for all
and
, it is eventually of one sign and hence
is eventually of one sign. So we shall distinguish the following two cases to discuss:
(I)
for
; and
(II)
for
.
Case (I). The proof that
is eventually positive is similar to that in Theorem 1, so it is omitted here.
Case (II). For
, we have
![]()
then
(3.11)
Integrating (3.11) from t (t ≥ T) to u (u ≥ t) and letting
, we have
![]()
and thus
(3.12)
where
. Applying (3.12) to Equation (1.1), we find
(3.13)
Integrating (3.13) from T to t, we have
![]()
Therefore,
(3.14)
Next integrating (3.14) from T to t, we obtain
![]()
By (3.9), we have
, which contradicts
. This completes the proof. □
Remark 2. By Theorem 2, we get the sufficient condition of oscillation for Equation (1.1) when the condition (1.3) is satisfied, while the usual result existing is that the conditions (1.3) was established, then every solution of the Equation (1.1) is either oscillatory or converges to zero on
.
Theorem 3. Based on (1.2), assume (H1)-(H4) hold and
. If there exists a positive D-
differentiable function
such that for every ![]()
(3.15)
where
is as the same as that in (3.2), then (1.1) is oscillatory on
.
Proof. Assume that (1.1) has a nonoscillatory solution x on
. Without loss of generality, we can assume that
and
for all
,
. By the definition of
, it follows
. The proof when x is eventually negative is similar. Proceeding as the proof of Theorem 1, we obtained (3.3) and (3.5). Using (3.3) in (3.5), we have that on ![]()
(3.16)
Also, since
, we have
![]()
i.e.,
(3.17)
Substituting (3.17) into (3.16), we obtain on ![]()
![]()
i.e.,
(3.18)
Now using inequality (3.7), we get
![]()
Hence, we have
![]()
This implies that on ![]()
(3.19)
Using (3.19) in (3.18), we have on
that
![]()
Integrating both sides of this inequality from T to t, taking the limsup of the resulting inequality as
and applying condition (3.15), we obtain a contradiction to the fact that
for
. This completes the proof. □
Using the same ideas as in the proof of Theorem 2, we can now obtain the following result based on (1.3).
Theorem 4. Under the condition (1.3), assume that the conditions (H1)-(H4), (3.9) and (3.15) hold, then (1.1) is oscillatory on
. □
4. Application
Now we shall reformulate the above conditions which are sufficient for the oscillation of (1.1) when (1.2) holds on different time scales:
If
, Equation (1.1) becomes
(4.1)
and then conditions (3.1) and (3.15), respectively, become
(4.2)
and
(4.3)
The conditions (4.2) and (4.3) are new.
If
, Equation (1.1) becomes
(4.4)
and conditions (3.1) and (3.15), respectively, become
(4.5)
and
(4.6)
At same time, the Theorems 1 and 3 are new for the case
.
Example 1. Consider the second-order nonlinear delay 2-difference equations
(4.7)
where
. This gives
![]()
The conditions (H1)-(H3) are clearly satisfied, and (H4) holds with L = 1. Because
![]()
(1.2) is satisfied. Now let
for all
, and then
![]()
Thus when
, we have
![]()
It is easy to see that (3.1) is satisfied as well. Altogether, the Equation (4.7) is oscillatory by Theorem 1.
Acknowledgements
We thank the Editor and the referee for their comments. Research of Q. Zhang is funded by the Natural Science Foundation of Shandong Province of China grant ZR2013AM003. This support is greatly appreciated.