Skip to main content

Oscillatory behavior of solutions of certain fractional difference equations

Abstract

In this paper, we consider the oscillation behavior of solutions of the following fractional difference equation:

$$ \Delta \bigl( c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \bigr) +q ( t ) G ( t ) =0, $$

where \(t\in \mathbf{N}_{t_{0}+1-\alpha }\), \(G ( t ) = \sum_{s=t_{0}}^{t-1+\alpha } ( t-s-1 ) ^{-\alpha }x ( s ) \), and \(\Delta^{\alpha }\) denotes a Riemann–Liouville fractional difference operator of order \(0<\alpha \leq 1\). By using the generalized Riccati transformation technique, we obtain some oscillation criteria. Finally we give an example.

Introduction and preliminaries

Fractional differential (or difference) equations are a more general form of differential equations with integer order. And there is an increasing interest in the study of them due to some important contributions [1, 2].

Many authors have been focused on various equations like ordinary and partial differential equations [3,4,5,6], difference equations [7,8,9], dynamic equations on time scales [10,11,12,13,14], and fractional differential (difference) equations [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31] obtaining some oscillation criteria. Recently, oscillation studies have become a very hot topic. That is why, we consider the following fractional difference equation:

$$ \Delta \bigl( c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \bigr) +q ( t ) G ( t ) =0, $$
(1)

where \(t\in \mathbf{N} _{t_{0}+1-\alpha }\), \(G ( t ) = \sum_{s=t_{0}}^{t-1+\alpha } ( t-s-1 ) ^{ ( -\alpha ) }x ( s ) \), \(c ( t ) \), \(a ( t ) \), \(r ( t ) \), and \(q ( t ) \) are positive sequences, and \(\Delta^{\alpha }\) denotes the Riemann–Liouville fractional difference operator of order \(0<\alpha \leq 1\).

By a solution of Eq. (1), we mean a real-valued sequence \(x ( t ) \) satisfying Eq. (1) for \(t\in \mathbf{N} _{t_{0}}\). A solution \(x ( t ) \) of Eq. (1) is called oscillatory if it is neither eventually positive nor eventually negative, otherwise it is called non-oscillatory. Equation (1) is called oscillatory if all its solutions are oscillatory.

Definition 1

([32])

Let \(v>0\). The vth fractional sum f is defined by

$$ \Delta^{-v}f ( t ) =\frac{1}{\varGamma ( v ) }\sum_{s=a} ^{t-v} ( t-s-1 ) ^{v-1}f ( s ) , $$
(2)

where f is defined for \(s\equiv a \mathbf{mod} ( 1 ) \), \(\Delta^{-v}f\) is defined for \(t\equiv ( a+v ) \mathbf{mod} ( 1 ) \), and \(t^{ ( v ) }=\frac{\varGamma ( t+1 ) }{\varGamma ( t-v+1 ) }\). The fractional sum \(\Delta^{-v}f\) maps functions defined on \(\mathbf{N} _{a}\) to functions defined on \(\mathbf{N} _{a+v}\), where \(\mathbf{N} _{t}= \{ t,t+1,t+2,\ldots \} \).

Definition 2

([32])

Let \(v>0\) and \(m-1<\mu <m\), where m denotes a positive integer, \(m= \lceil \mu \rceil \). Set \(v=m-\mu \). The μth fractional difference is defined as

$$ \Delta^{\mu }f ( t ) =\Delta^{m-v}f ( t ) =\Delta^{m} \Delta^{-v}f ( t ) , $$
(3)

where \(\lceil \mu \rceil \) is the ceiling function of μ.

Lemma 1

([33])

Assume that A and B are nonnegative real numbers. Then

$$ \lambda AB^{\lambda -1}-A^{\lambda }\leq ( \lambda -1 ) B ^{\lambda } $$
(4)

for all \(\lambda >1\).

Main results

Throughout this paper, we denote

$$ \phi ( t ) =\sum_{s=t_{1}}^{t-1} \frac{1}{c ( s ) };\quad\quad \vartheta ( t ) =\sum_{s=t_{2}}^{t-1} \frac{\phi ( s ) }{a ( s ) };\quad\quad \delta ( t ) =\sum_{s=t_{3}} ^{t-1}\frac{\vartheta ( s ) }{r ( s ) }. $$

For simplification, we consider

$$ \Delta \gamma_{+} ( s ) =\max \bigl\{ 0,\Delta \gamma ( s ) \bigr\} $$

and

$$ \Delta \beta_{+} ( s ) =\max \bigl\{ 0,\Delta \beta ( s ) \bigr\} . $$

Lemma 2

([28])

Let \(x ( t ) \) be a solution of Eq. (1), and let

$$ G ( t ) =\sum_{s=t_{0}}^{t-1+\alpha } ( t-s-1 ) ^{ ( -\alpha ) }x ( s ) , $$
(5)

then

$$ \Delta \bigl( G ( t ) \bigr) =\varGamma ( 1-\alpha ) \Delta^{\alpha }x ( t ) . $$
(6)

Lemma 3

Assume that \(x ( t ) \) is an eventually positive solution of Eq. (1). If

$$ \sum_{s=t_{0}}^{\infty }\frac{1}{c ( s ) }=\sum _{s=t_{0}} ^{\infty }\frac{1}{a ( s ) }=\sum _{s=t_{0}}^{\infty }\frac{1}{r ( s ) }=\infty , $$
(7)

then we have two possible cases for \(t\in [ t_{1},\infty ) \), \(t_{1}>t_{0}\) is sufficiently large:

  1. Case 1

    \(\Delta^{\alpha }x ( t ) >0\), \(\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) >0\), \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) >0\) or

  2. Case 2

    \(\Delta^{\alpha }x ( t ) >0\), \(\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) <0\), \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) >0\).

Proof

From the hypothesis, there exists \(t_{1}\) such that \(x ( t ) >0\) on \([ t_{1},\infty ) \), so that \(G ( t ) >0\) on \([ t_{1},\infty ) \), and from Eq. (1), we have

$$ \Delta \bigl( c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \bigr) =-q ( t ) G ( t ) < 0. $$
(8)

Then \(c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \) is an eventually non-increasing sequence on \([ t_{1},\infty ) \). We know that \(\Delta^{\alpha }x ( t ) \), \(\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) \), and \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \) are eventually of one sign. For \(t_{2}>t_{1}\) is sufficiently large, we claim that \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) >0\) on \([ t_{2}, \infty ) \). Otherwise, assume that there exists sufficiently large \(t_{3}>t_{2}\) such that \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) <0\) on \([ t_{3},\infty ) \). For \([ t_{3},\infty ) \) and there exists a constant \(l_{1}>0\), we have

$$ \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \leq -\frac{l_{1}}{c ( t ) }< 0. $$

Hence, there exist a constant \(l_{2}>0\) and sufficiently large \(t_{4}>t_{3}\) such that

$$ \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \leq - \frac{l_{2}}{a ( t ) }< 0. $$
(9)

Then there exist a constant \(l_{3}>0\) and sufficiently large \(t_{5}>t_{4}\) such that

$$ \Delta^{\alpha }x ( t ) \leq -\frac{l_{3}}{r ( t ) }, $$

that is,

$$ \Delta G ( t ) \leq -\frac{\varGamma ( 1-\alpha ) l _{3}}{r ( t ) }< 0. $$

By (7), we obtain \(\lim_{t\rightarrow \infty }G ( t ) =-\infty \). This is a contradiction. If \(\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) <0\), then \(\Delta^{\alpha }x ( t ) >0\) due to \(\sum_{s=t_{0}}^{\infty }\frac{1}{r ( s ) }=\infty \). If \(\Delta ( r ( t ) \Delta^{ \alpha }x ( t ) ) >0\), then \(\Delta^{\alpha }x ( t ) >0\) due to \(\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) >0\). So, the proof is complete. □

Lemma 4

Assume that \(x ( t ) \) is an eventually positive solution of Eq. (1), which satisfies Case 1 of Lemma 3. Then

$$ a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \geq c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }. $$

If there exists a positive sequence ϕ such that, for \(t\in [ t_{1},\infty ) \),

$$ \frac{\phi ( t ) }{c ( t ) \sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }}-\Delta \phi ( t ) \leq 0, $$

where \(t_{1}\) is sufficiently large, then \(a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) / \phi ( t ) \) is a non-increasing sequence on \([ t_{1}, \infty ) \) and

$$ r ( t ) \Delta^{\alpha }x ( t ) \geq \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \frac{a ( t ) }{\phi ( t ) }\sum _{s=t_{1}}^{t-1}\frac{\phi ( s ) }{a ( s ) }. $$

Furthermore, if there exists a positive sequence ϑ and \(t_{2}>t_{1}\) is sufficiently large such that, for \(t\in [ t_{2}, \infty ) \),

$$ \frac{\vartheta ( t ) }{\frac{a ( t ) }{\phi ( t ) }\sum_{s=t_{2}}^{t-1}\frac{\phi ( s ) }{a ( s ) }}-\Delta \vartheta ( t ) \leq 0, $$

then \(r ( t ) \Delta^{\alpha }x ( t ) /\vartheta ( t ) \) is a non-increasing sequence on \([ t_{2}, \infty ) \) and

$$ G ( t ) \geq \Delta G ( t ) \frac{r ( t ) }{ \vartheta ( t ) }\sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) }. $$

Suppose also that there exists a positive sequence δ and \(t_{3}>t_{2}\) is sufficiently large such that, for \(t\in [ t_{3}, \infty ) \),

$$ \frac{\delta ( t ) }{\frac{r ( t ) }{\vartheta ( t ) }\sum_{s=t_{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) }}-\Delta \delta ( t ) \leq 0. $$

Then \(G ( t ) /\delta ( t ) \) is a non-increasing sequence on \([ t_{3},\infty ) \).

Proof

Assume that x is an eventually positive solution of Eq. (1). Then we have that \(\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) >0\) and \(\Delta ( c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) ) <0\) on \([ t_{0},\infty ) \). So,

$$\begin{aligned} a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) =&a ( t_{0} ) \Delta \bigl( r ( t_{0} ) \Delta^{\alpha }x ( t_{0} ) \bigr) \\ &{}+\sum_{s=t_{0}}^{t-1}\frac{c ( s ) \Delta ( a ( s ) \Delta ( r ( s ) \Delta^{\alpha }x ( s ) ) ) }{c ( s ) } \\ \geq & c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \sum_{s=t_{0}}^{t-1} \frac{1}{c ( s ) }, \end{aligned}$$

and then

$$\begin{aligned}& \Delta \biggl( \frac{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{\phi ( t ) } \biggr) \\& \quad = \frac{\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \phi ( t ) -a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) \Delta \phi ( t ) }{\phi ( t ) \phi ( t+1 ) } \\& \quad \leq \frac{\Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{\phi ( t ) \phi ( t+1 ) } \biggl( \frac{\phi ( t ) }{c ( t ) \sum_{s=t_{1}}^{t-1}\frac{1}{c ( s ) }}-\Delta \phi ( t ) \biggr) \leq 0. \end{aligned}$$

Hence, \(a ( t ) \Delta ( r ( t ) \Delta^{ \alpha }x ( t ) ) /\phi ( t ) \) is a non-increasing sequence on \([ t_{1},\infty ) \) where \(t_{1}>t_{0}\) is sufficiently large. Then we have

$$\begin{aligned} r ( t ) \Delta^{\alpha }x ( t ) =&r ( t_{1} ) \Delta^{\alpha }x ( t_{1} ) +\sum_{s=t_{1}}^{t-1} \frac{a ( s ) \Delta ( r ( s ) \Delta^{\alpha }x ( s ) ) }{\phi ( s ) }\frac{\phi ( s ) }{a ( s ) } \\ \geq &\frac{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{\phi ( t ) } \sum_{s=t_{1}}^{t-1} \frac{\phi ( s ) }{a ( s ) } \end{aligned}$$

and

$$\begin{aligned} \Delta \biggl( \frac{r ( t ) \Delta^{\alpha }x ( t ) }{\vartheta ( t ) } \biggr) =&\frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) \vartheta ( t ) -r ( t ) \Delta^{\alpha }x ( t ) \Delta \vartheta ( t ) }{\vartheta ( t ) \vartheta ( t+1 ) } \\ \leq &\frac{r ( t ) \Delta^{\alpha }x ( t ) }{ \vartheta ( t ) \vartheta ( t+1 ) } \biggl( \frac{ \vartheta ( t ) }{\frac{a ( t ) }{\phi ( t ) }\sum_{s=t_{1}}^{t-1}\frac{\phi ( s ) }{a ( s ) }}-\Delta \vartheta ( t ) \biggr) \leq 0. \end{aligned}$$

So \(r ( t ) \Delta^{\alpha }x ( t ) /\vartheta ( t ) \) is a non-increasing sequence on \([ t_{2}, \infty ) \) where \(t_{2}>t_{1}\) is sufficiently large. Then we have

$$\begin{aligned} G ( t ) =&G ( t_{2} ) +\varGamma ( 1-\alpha ) \sum _{s=t_{2}}^{t-1}\frac{r ( s ) \Delta^{\alpha }x ( s ) }{\vartheta ( s ) }\frac{ \vartheta ( s ) }{r ( s ) } \\ \geq &\frac{r ( t ) \varGamma ( 1-\alpha ) \Delta^{\alpha }x ( t ) }{\vartheta ( t ) } \sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) } \\ =&\Delta G ( t ) \frac{r ( t ) }{\vartheta ( t ) }\sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) }, \end{aligned}$$

and then

$$\begin{aligned} \Delta \biggl( \frac{G ( t ) }{\delta ( t ) } \biggr) =&\frac{ ( \Delta G ( t ) ) \delta ( t ) -G ( t ) \Delta \delta ( t ) }{\delta ( t ) \delta ( t+1 ) } \\ \leq &\frac{G ( t ) }{\delta ( t ) \delta ( t+1 ) } \biggl( \frac{\delta ( t ) }{\frac{r ( t ) }{\vartheta ( t ) }\sum_{s=t_{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) }}-\Delta \delta ( t ) \biggr) \leq 0. \end{aligned}$$

Then \(G ( t ) /\delta ( t ) \) is a non-increasing sequence on \([ t_{3},\infty ) \) where \(t_{3}>t_{2}\) is sufficiently large. So the proof is complete. □

Theorem 1

Assume that (7) holds and there exists a positive sequence γ such that, for all sufficiently large t,

$$ \lim_{t\rightarrow \infty }\sup \sum_{s=t_{3}}^{t-1} \Biggl( \frac{ \varGamma ( 1-\alpha ) \gamma ( s ) q ( s ) }{\vartheta ( s ) \phi ( s+1 ) }\sum_{u=t_{2}} ^{s-1} \frac{\vartheta ( u ) }{r ( u ) }\sum_{u=t _{1}}^{s-1} \frac{\phi ( u ) }{a ( u ) }-\frac{c ( s ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4 \gamma ( s ) } \Biggr) =\infty . $$
(10)

If there exist positive sequences β, λ such that, for all sufficiently large t,

$$ \frac{\lambda ( t ) }{r ( t ) \sum_{s=t_{1}}^{t-1}\frac{1}{r ( s ) }}-\Delta \lambda ( t ) \leq 0 $$
(11)

and

$$ \lim_{t\rightarrow \infty }\sup \sum_{\zeta =t_{2}}^{t-1} \Biggl( \frac{ \beta ( \zeta ) \lambda ( \zeta ) }{\lambda ( \zeta +1 ) a ( \zeta ) }\sum_{s=\zeta }^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s}^{\infty }q ( v ) \Biggr) -\frac{r ( \zeta ) ( \Delta \beta_{+} ( \zeta ) ) ^{2}}{4\varGamma ( 1-\alpha ) \beta ( \zeta ) } \Biggr) =\infty . $$
(12)

Then every solution of Eq. (1) is oscillatory.

Proof

Suppose to the contrary that \(x(t)\) is a non-oscillatory solution of Eq. (1). Then, without loss of generality, we may assume that there is a solution \(x ( t ) \) of Eq. (1) such that \(x ( t ) >0\) on \([ t_{0},\infty ) \), where \(t_{0}\) is sufficiently large. From Lemma 3, \(x ( t ) \) satisfies Case 1 or Case 2. Firstly, let Case 1 hold. Then we define the following function:

$$ \omega ( t ) =\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }. $$

For \(t\in [ t_{0},\infty ) \), we have

$$\begin{aligned} \Delta \omega ( t ) =&\Delta \gamma ( t ) \frac{ \omega ( t+1 ) }{\gamma ( t+1 ) }+\gamma ( t ) \Delta \biggl( \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) } \biggr) \\ =&\Delta \gamma ( t ) \frac{\omega ( t+1 ) }{ \gamma ( t+1 ) }-\gamma ( t ) \frac{q ( t ) G ( t ) }{a ( t+1 ) \Delta ( r ( t+1 ) \Delta^{\alpha }x ( t+1 ) ) } \\ &{}-\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) a ( t+1 ) \Delta ( r ( t+1 ) \Delta^{\alpha }x ( t+1 ) ) }. \end{aligned}$$

Since \(a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) /\phi ( t ) \) is a non-increasing sequence on \([ t_{1},\infty ) \), we have

$$ \frac{a ( t+1 ) \Delta ( r ( t+1 ) \Delta^{ \alpha }x ( t+1 ) ) }{\phi ( t+1 ) }\leq \frac{a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{\phi ( t ) }. $$

From Lemma 4, we obtain

$$\begin{aligned}& \frac{G ( t ) }{a ( t+1 ) \Delta ( r ( t+1 ) \Delta^{\alpha }x ( t+1 ) ) } \\& \quad = \frac{1}{a ( t+1 ) }\frac{G ( t ) }{\Delta G ( t ) }\frac{\Delta G ( t ) }{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }\frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{\Delta ( r ( t+1 ) \Delta^{\alpha }x ( t+1 ) ) } \\& \quad \geq \frac{1}{a ( t+1 ) } \Biggl( \frac{r ( t ) }{ \vartheta ( t ) }\sum _{s=t_{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) } \Biggr) \Biggl( \frac{\varGamma ( 1- \alpha ) }{r ( t ) }\frac{a ( t ) }{\phi ( t ) }\sum_{s=t_{1}}^{t-1} \frac{\phi ( s ) }{a ( s ) } \Biggr) \frac{\phi ( t ) a ( t+1 ) }{\phi ( t+1 ) a ( t ) } \\& \quad = \frac{\varGamma ( 1-\alpha ) }{\vartheta ( t ) \phi ( t+1 ) }\sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) } \Biggl( \sum_{s=t_{1}}^{t-1} \frac{\phi ( s ) }{a ( s ) } \Biggr) \end{aligned}$$

and

$$\begin{aligned} \Delta \omega ( t ) \leq &\Delta \gamma_{+} ( t ) \frac{\omega ( t+1 ) }{\gamma ( t+1 ) }-\gamma ( t ) q ( t ) \frac{\varGamma ( 1-\alpha ) }{\vartheta ( t ) \phi ( t+1 ) }\sum _{s=t_{2}} ^{t-1}\frac{\vartheta ( s ) }{r ( s ) } \Biggl( \sum _{s=t_{1}}^{t-1}\frac{\phi ( s ) }{a ( s ) } \Biggr) \\ &{}-\frac{\gamma ( t ) }{c ( t ) }\frac{\omega^{2} ( t+1 ) }{\gamma^{2} ( t+1 ) }. \end{aligned}$$

Setting \(\lambda =2\), \(A= ( \frac{\gamma ( t ) }{c ( t ) } ) ^{1/2}\frac{\omega ( t+1 ) }{\phi ( t+1 ) }\), and \(B=\frac{1}{2} ( \frac{c ( t ) }{ \gamma ( t ) } ) ^{1/2}\Delta \gamma_{+} ( t ) \) using Lemma 1, we obtain

$$ \Delta \omega ( t ) \leq -\gamma ( t ) q ( t ) \frac{\varGamma ( 1-\alpha ) }{\vartheta ( t ) \phi ( t+1 ) }\sum _{s=t_{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) } \Biggl( \sum _{s=t_{1}}^{t-1}\frac{ \phi ( s ) }{a ( s ) } \Biggr) + \frac{c ( t ) }{4\gamma ( t ) } \bigl( \Delta \gamma_{+} ( t ) \bigr) ^{2}. $$

Summing both sides of the above inequality from \(t_{3}\) to \(t-1\), we get

$$\begin{aligned}& \sum_{s=t_{3}}^{t-1} \Biggl( \frac{\varGamma ( 1-\alpha ) \gamma ( s ) q ( s ) }{\vartheta ( s ) \phi ( s+1 ) } \sum_{u=t_{2}}^{s-1} \frac{\vartheta ( u ) }{r ( u ) } \Biggl( \sum_{u=t_{1}}^{s-1} \frac{\phi ( u ) }{a ( u ) } \Biggr) - \frac{c ( s ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4\gamma ( s ) } \Biggr) \\& \quad \leq \omega ( t_{3} ) -\omega ( t ) \leq \omega ( t_{3} ) . \end{aligned}$$

This contradicts (10). Now we consider Case 2. Then we define the following function:

$$ \omega_{2} ( t ) =\beta ( t ) \frac{r ( t ) \Delta^{\alpha }x ( t ) }{G ( t ) }. $$

Then

$$\begin{aligned} \Delta \omega_{2} ( t ) =&\Delta \beta ( t ) \frac{ \omega ( t+1 ) }{\beta ( t+1 ) }+ \beta ( t ) \Delta \biggl( \frac{r ( t ) \Delta^{\alpha }x ( t ) }{G ( t ) } \biggr) \\ =&\Delta \beta ( t ) \frac{\omega ( t+1 ) }{ \beta ( t+1 ) }+\beta ( t ) \biggl( \frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) G ( t ) -r ( t ) \Delta^{\alpha }x ( t ) \Delta G ( t ) }{G ( t ) G ( t+1 ) } \biggr) \\ =&\Delta \beta ( t ) \frac{\omega ( t+1 ) }{ \beta ( t+1 ) }+\beta ( t ) \frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{G ( t+1 ) }-\beta ( t ) \frac{r ( t ) \Delta^{ \alpha }x ( t ) \Delta G ( t ) }{G ( t ) G ( t+1 ) }. \end{aligned}$$

Hence we have

$$\begin{aligned} G ( t ) =&G ( t_{1} ) +\varGamma ( 1-\alpha ) \sum _{s=t_{1}}^{t-1}\frac{r ( s ) \Delta^{\alpha }x ( s ) }{r ( s ) } \\ \geq &\varGamma ( 1-\alpha ) r ( t ) \Delta^{ \alpha }x ( t ) \sum _{s=t_{1}}^{t-1}\frac{1}{r ( s ) }. \end{aligned}$$

That is,

$$ \frac{G ( t ) }{r ( t ) \sum_{s=t_{1}}^{t-1}\frac{1}{r ( s ) }}\geq \varGamma ( 1-\alpha ) \Delta^{\alpha }x ( t ) = \Delta G ( t ) $$

and

$$\begin{aligned} \Delta \biggl( \frac{G ( t ) }{\lambda ( t ) } \biggr) =&\frac{\Delta G ( t ) \lambda ( t ) -G ( t ) \Delta \lambda ( t ) }{\lambda ( t ) \lambda ( t+1 ) } \\ \leq &\frac{G ( t ) }{\lambda ( t ) \lambda ( t+1 ) } \biggl( \frac{\lambda ( t ) }{r ( t ) \sum_{s=t_{1}}^{t-1}\frac{1}{r ( s ) }}-\Delta \lambda ( t ) \biggr) \leq 0. \end{aligned}$$

Thus we have \(G ( t ) /\lambda ( t ) \) is eventually non-increasing and

$$ \frac{G ( t ) }{G ( t+1 ) }\geq \frac{\lambda ( t ) }{\lambda ( t+1 ) }. $$
(13)

Using the fact that \(r ( t ) \Delta^{\alpha }x ( t ) \) is strictly decreasing, we have

$$ r ( t ) \Delta^{\alpha }x ( t ) \geq r ( t+1 ) \Delta^{\alpha }x ( t+1 ) $$

and \(\Delta G ( t ) >0\), then \(G ( t+1 ) >G ( t ) \), it follows that

$$\begin{aligned} \Delta \omega_{2} ( t ) \leq &\Delta \beta_{+} ( t ) \frac{\omega ( t+1 ) }{\beta ( t+1 ) }+\beta ( t ) \frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{G ( t+1 ) } \\ &{}-\frac{\varGamma ( 1-\alpha ) \beta ( t ) }{r ( t ) }\frac{\omega_{2}^{2} ( t+1 ) }{\beta^{2} ( t+1 ) }. \end{aligned}$$

From 8, we have

$$\begin{aligned}& c ( u ) \Delta \bigl( a ( u ) \Delta \bigl( r ( u ) \Delta^{\alpha }x ( u ) \bigr) \bigr) -c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \\ & \quad =-\sum_{s=t}^{u-1}q ( s ) G ( s ) \end{aligned}$$

for \(\Delta G ( t ) >0\), and letting \(u\rightarrow \infty \), we get

$$ -c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \leq -G ( t ) \sum_{s=t}^{\infty }q ( s ) $$

or

$$ \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) \geq \frac{G ( t ) }{c ( t ) }\sum_{s=t}^{\infty }q ( s ) . $$

And so

$$ a ( u ) \Delta \bigl( r ( u ) \Delta^{\alpha }x ( u ) \bigr) -a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \geq G ( t ) \sum _{s=t} ^{u-1} \Biggl( \frac{1}{c ( s ) }\sum _{v=s}^{\infty }q ( v ) \Biggr) . $$

Letting \(u\rightarrow \infty \), we have

$$ \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \leq -G ( t ) \frac{1}{a ( t ) }\sum_{s=t}^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s}^{\infty }q ( v ) \Biggr) $$

due to \(\lim_{u\rightarrow \infty }a ( u ) \Delta ( r ( u ) \Delta^{\alpha }x ( u ) ) =k<0\). Then, by (13), we obtain

$$\begin{aligned} \frac{\Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{G ( t+1 ) } \leq& -\frac{G ( t ) }{G ( t+1 ) }\frac{1}{a ( t ) }\sum _{s=t}^{\infty } \Biggl( \frac{1}{c ( s ) }\sum _{v=s}^{\infty }q ( v ) \Biggr) \\ \leq &-\frac{\lambda ( t ) }{\lambda ( t+1 ) }\frac{1}{a ( t ) }\sum _{s=t}^{\infty } \Biggl( \frac{1}{c ( s ) }\sum _{v=s}^{\infty }q ( v ) \Biggr) . \end{aligned}$$

So,

$$\begin{aligned} \begin{aligned} \Delta \omega_{2} ( t ) &\leq \Delta \beta_{+} ( t ) \frac{\omega_{2} ( t+1 ) }{\beta ( t+1 ) }-\beta ( t ) \frac{\lambda ( t ) }{\lambda ( t+1 ) } \frac{1}{a ( t ) }\sum_{s=t}^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s}^{\infty }q ( v ) \Biggr) \\ &{}-\frac{\varGamma ( 1-\alpha ) \beta ( t ) }{r ( t ) }\frac{\omega_{2}^{2} ( t+1 ) }{\beta^{2} ( t+1 ) }.\end{aligned} \end{aligned}$$

Setting \(\lambda =2\), \(A= ( \frac{\varGamma ( 1-\alpha ) \beta ( t ) }{r ( t ) } ) ^{1/2}\frac{\omega_{2} ( t+1 ) }{\beta ( t+1 ) }\), and \(B=\frac{1}{2} ( \frac{r ( t ) }{\varGamma ( 1-\alpha ) \beta ( t ) } ) ^{1/2}\Delta \beta_{+} ( t ) \) using Lemma 1, we obtain

$$ \Delta \omega_{2} ( t ) \leq -\beta ( t ) \frac{ \lambda ( t ) }{\lambda ( t+1 ) } \frac{1}{a ( t ) }\sum_{s=t}^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s} ^{\infty }q ( v ) \Biggr) +\frac{r ( t ) ( \Delta \beta_{+} ( t ) ) ^{2}}{4\varGamma ( 1-\alpha ) \beta ( t ) }. $$

Summing both sides of the above inequality from \(t_{2}\) to \(t-1\), we have

$$\begin{aligned}& \sum_{\zeta =t_{2}}^{t-1} \Biggl( \beta ( \zeta ) \frac{ \lambda ( \zeta ) }{\lambda ( \zeta +1 ) }\frac{1}{a ( \zeta ) }\sum_{s=\zeta }^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s}^{\infty }q ( v ) \Biggr) -\frac{r ( \zeta ) ( \Delta \beta_{+} ( \zeta ) ) ^{2}}{4\varGamma ( 1-\alpha ) \beta ( \zeta ) } \Biggr) \\& \quad \leq \omega_{2} ( t_{2} ) -\omega_{2} ( t ) \leq \omega_{2} ( t_{2} ) < \infty , \end{aligned}$$

which contradicts (12). So, the proof is complete. □

Theorem 2

Let (7) hold. Assume that there exists a positive sequence γ such that, for all sufficiently large t,

$$ \lim_{t\rightarrow \infty }\sup \sum_{s=t_{3}}^{t-1} \Biggl( \gamma ( s ) q ( s ) \frac{\varGamma ( 1-\alpha ) }{\vartheta ( s+1 ) }\sum _{u=t_{2}}^{s-1}\frac{\vartheta ( u ) }{r ( u ) }-\frac{a ( s ) \vartheta ( s+1 ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4\gamma ( s ) \vartheta ( s ) \sum_{u=t_{0}} ^{s-1}\frac{1}{c ( u ) }} \Biggr) =\infty . $$
(14)

If there exist positive sequences β, λ such that (11) and (12) hold, then Eq. (1) is oscillatory.

Proof

Suppose to the contrary that \(x(t)\) is a non-oscillatory solution of (1). Then, without loss of generality, we may assume that there is a solution \(x ( t ) \) of Eq. (1) such that \(x ( t ) >0\) on \([ t_{0},\infty ) \) where \(t_{0}\) is sufficiently large. From Lemma 3, \(x ( t ) \) satisfies Case 1 or Case 2. Firstly, let Case 1 hold. Then we define the following function:

$$ \pi ( t ) =\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{r ( t ) \Delta^{\alpha }x ( t ) }. $$

For \(t\in [ t_{0},\infty ) \), we have

$$\begin{aligned} \Delta \pi ( t ) =&\Delta \gamma ( t ) \frac{ \pi ( t+1 ) }{\gamma ( t+1 ) }+\gamma ( t ) \Delta \biggl( \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{r ( t ) \Delta^{\alpha }x ( t ) } \biggr) \\ =&\Delta \gamma ( t ) \frac{\pi ( t+1 ) }{ \gamma ( t+1 ) }-\gamma ( t ) \frac{q ( t ) G ( t ) }{r ( t+1 ) \Delta^{\alpha }x ( t+1 ) } \\ &{}-\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) }{r ( t ) \Delta^{\alpha }x ( t ) r ( t+1 ) \Delta^{\alpha }x ( t+1 ) }. \end{aligned}$$

From Lemma 4, we obtain

$$\begin{aligned}& \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \geq \frac{\sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }}{a ( t ) }c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) , \\& 1\leq \frac{r ( t+1 ) \Delta^{\alpha }x ( t+1 ) }{r ( t ) \Delta^{\alpha }x ( t ) }\leq \frac{\vartheta ( t+1 ) }{\vartheta ( t ) }, \\& \frac{\vartheta ( t ) }{\vartheta ( t+1 ) }\leq \frac{r ( t+1 ) \Delta^{\alpha }x ( t+1 ) }{r ( t ) \Delta^{\alpha }x ( t ) } \end{aligned}$$

or

$$ \frac{r ( t+1 ) \vartheta ( t ) }{r ( t ) \vartheta ( t+1 ) }\leq \frac{\Delta G ( t ) }{ \Delta G ( t+1 ) } $$

and

$$\begin{aligned} \frac{G ( t ) }{r ( t+1 ) \Delta^{\alpha }x ( t+1 ) } =&\frac{\varGamma ( 1-\alpha ) }{r ( t+1 ) }\frac{G ( t ) }{\Delta G ( t ) } \frac{ \Delta G ( t ) }{\Delta G ( t+1 ) } \\ \geq &\frac{\varGamma ( 1-\alpha ) }{r ( t+1 ) } \Biggl( \frac{r ( t ) }{\vartheta ( t ) }\sum _{s=t _{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) } \Biggr) \frac{r ( t+1 ) \vartheta ( t ) }{r ( t ) \vartheta ( t+1 ) } \\ =&\frac{\varGamma ( 1-\alpha ) }{\vartheta ( t+1 ) }\sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) }. \end{aligned}$$

Hence,

$$\begin{aligned} \Delta \pi ( t ) \leq &\Delta \gamma_{+} ( t ) \frac{ \pi ( t+1 ) }{\gamma ( t+1 ) }- \gamma ( t ) q ( t ) \frac{\varGamma ( 1-\alpha ) }{ \vartheta ( t+1 ) }\sum_{s=t_{2}}^{t-1} \frac{\vartheta ( s ) }{r ( s ) } \\ &{}-\frac{\gamma ( t ) \vartheta ( t ) }{\vartheta ( t+1 ) }\frac{\sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }}{a ( t ) }\frac{\pi^{2} ( t+1 ) }{\gamma^{2} ( t+1 ) }. \end{aligned}$$

In Lemma 1, choosing \(\lambda =2\), \(A= ( \frac{\gamma ( t ) \vartheta ( t ) }{\vartheta ( t+1 ) }\frac{ \sum_{s=t_{1}}^{t-1}\frac{1}{c ( s ) }}{a ( t ) } ) ^{1/2}\frac{\pi ( t+1 ) }{\gamma ( t+1 ) }\), and \(B=\frac{1}{2} ( \frac{a ( t ) \vartheta ( t+1 ) }{\gamma ( t ) \vartheta ( t ) \sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }} ) ^{1/2} \Delta \gamma_{+} ( t ) \), we obtain

$$ \Delta \pi ( t ) \leq -\gamma ( t ) q ( t ) \frac{\varGamma ( 1-\alpha ) }{\vartheta ( t+1 ) } \sum _{s=t_{2}}^{t-1}\frac{\vartheta ( s ) }{r ( s ) }+\frac{a ( t ) \vartheta ( t+1 ) ( \Delta \gamma_{+} ( t ) ) ^{2}}{4\gamma ( t ) \vartheta ( t ) \sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }}. $$

Summing both sides of the above inequality from \(t_{3}\) to \(t-1\), we have

$$\begin{aligned}& \sum_{s=t_{3}}^{t-1} \Biggl( \gamma ( s ) q ( s ) \frac{ \varGamma ( 1-\alpha ) }{\vartheta ( s+1 ) } \sum_{u=t_{2}}^{s-1} \frac{\vartheta ( u ) }{r ( u ) }-\frac{a ( s ) \vartheta ( s+1 ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4\gamma ( s ) \vartheta ( s ) \sum_{u=t_{0}}^{s-1}\frac{1}{c ( u ) }} \Biggr) \\& \quad \leq \pi ( t_{1} ) -\pi ( t ) \\& \quad \leq \pi ( t_{2} ) < \infty , \end{aligned}$$

which contradicts (14). And the proof of Case 2 is the same as that of Theorem 1 and hence is omitted. This completes the proof. □

Theorem 3

Let (7) hold. Assume that there exists a positive sequence γ such that, for all sufficiently large t,

$$ \lim_{t\rightarrow \infty }\sup \sum_{s=t_{2}}^{t-1} \biggl( \gamma ( s ) q ( s ) \frac{\delta ( s ) }{ \delta ( s+1 ) }-\frac{r ( s ) \phi ( s ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4\gamma ( s ) \sum_{s=t_{1}}^{u-1}\frac{\phi ( u ) }{a ( u ) }\sum_{u=t_{0}}^{s-1}\frac{1}{c ( u ) }} \biggr) = \infty . $$
(15)

If there exist positive sequences β, λ such that (11) and (12) hold, then Eq. (1) is oscillatory.

Proof

Suppose to the contrary that \(x(t)\) is a non-oscillatory solution of (1). Then, without loss of generality, we may assume that there is a solution \(x ( t ) \) of Eq. (1) such that \(x ( t ) >0\) on \([ t_{0},\infty ) \), where \(t_{0}\) is sufficiently large. From Lemma 3, \(x ( t ) \) satisfies Case 1 or Case 2. Firstly, let Case 1 hold. Then we define the following function:

$$ \nu ( t ) =\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{G ( t ) }. $$

For \(t\in [ t_{0},\infty ) \), we get

$$\begin{aligned} \Delta \nu ( t ) =&\Delta \gamma ( t ) \frac{ \nu ( t+1 ) }{\gamma ( t+1 ) }+\gamma ( t ) \Delta \biggl( \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) }{G ( t ) } \biggr) \\ =&\Delta \gamma ( t ) \frac{\nu ( t+1 ) }{ \alpha ( t+1 ) }-\gamma ( t ) \frac{q ( t ) G ( t ) }{G ( t+1 ) } \\ &{}-\gamma ( t ) \frac{c ( t ) \Delta ( a ( t ) \Delta ( r ( t ) \Delta^{\alpha }x ( t ) ) ) \Delta G ( t ) }{G ( t ) G ( t+1 ) }. \end{aligned}$$

From Lemma 4, we have

$$ \Delta G ( t ) \geq \frac{1}{r ( t ) } \Biggl( \frac{a ( t ) }{\phi ( t ) }\sum _{s=t_{1}}^{t-1}\frac{ \phi ( s ) }{a ( s ) } \Biggr) \frac{\sum_{s=t_{0}} ^{t-1}\frac{1}{c ( s ) }}{a ( t ) }c ( t ) \Delta \bigl( a ( t ) \Delta \bigl( r ( t ) \Delta^{\alpha }x ( t ) \bigr) \bigr) $$

and

$$ \frac{G ( t ) }{G ( t+1 ) }\geq \frac{\delta ( t ) }{\delta ( t+1 ) }. $$

Thus we obtain

$$\begin{aligned} \Delta \nu ( t ) \leq &\Delta \gamma_{+} ( t ) \frac{ \nu ( t+1 ) }{\gamma ( t+1 ) }- \gamma ( t ) p ( t ) \frac{\delta ( t ) }{\delta ( t+1 ) } \\ &{}-\frac{\gamma ( t ) }{r ( t ) \phi ( t ) }\sum_{s=t_{1}}^{t-1} \frac{\phi ( s ) }{a ( s ) } \sum_{s=t_{0}}^{t-1} \frac{1}{c ( s ) }\frac{\nu^{2} ( t+1 ) }{\gamma^{2} ( t+1 ) }. \end{aligned}$$

Then, setting \(\lambda =2\),

$$\begin{aligned}& A= \Biggl( \frac{\gamma ( t ) }{r ( t ) \phi ( t ) }\sum_{s=t_{1}}^{t-1} \frac{ \phi ( s ) }{a ( s ) }\sum_{s=t_{0}}^{t-1} \frac{1}{c ( s ) } \Biggr) ^{1/2}\frac{\nu ( t+1 ) }{\gamma ( t+1 ) },\quad \text{and} \\& B=\frac{1}{2} \biggl( \frac{r ( t ) \phi ( t ) }{\gamma ( t ) \sum_{s=t_{1}}^{t-1}\frac{ \phi ( s ) }{a ( s ) }\sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }} \biggr) ^{1/2}\Delta \gamma_{+} ( t ) \end{aligned}$$

using Lemma 1, we obtain

$$ \Delta \nu ( t ) \leq -\gamma ( t ) q ( t ) \frac{\delta ( t ) }{\delta ( t+1 ) }+ \frac{r ( t ) \phi ( t ) ( \Delta \gamma_{+} ( t ) ) ^{2}}{4\gamma ( t ) \sum_{s=t_{1}}^{t-1}\frac{ \phi ( s ) }{a ( s ) }\sum_{s=t_{0}}^{t-1}\frac{1}{c ( s ) }}. $$

Summing both sides of the above inequality from \(t_{2}\) to \(t-1\), we have

$$\begin{aligned} \begin{aligned} \sum_{s=t_{2}}^{t-1} \biggl( \gamma ( s ) q ( s ) \frac{ \delta ( s ) }{\delta ( s+1 ) }-\frac{r ( s ) \phi ( s ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4\gamma ( s ) \sum_{s=t_{1}}^{u-1}\frac{ \phi ( u ) }{a ( u ) }\sum_{u=t_{0}}^{s-1}\frac{1}{c ( u ) }} \biggr) & \leq \nu ( t_{2} ) -\nu ( t ) \\ & \leq \nu ( t_{2} ) < \infty ,\end{aligned} \end{aligned}$$

which contradicts (15). The proof of Case 2 is the same as that of Theorem 1 and hence is omitted. This completes the proof. □

Applications

Example 1

Consider the following fractional difference equation for \(t\geq 2\):

$$ \Delta^{3+\alpha }x ( t ) +t^{-2} \Biggl( \sum _{s=t_{0}}^{t-1+ \alpha } ( t-s-1 ) ^{ ( -\alpha ) }x ( s ) \Biggr) =0. $$
(16)

This corresponds to Eq. (1) with \(\alpha \in ( 0,1 ] \), \(t_{0}=2\), \(c ( t ) =a ( t ) =r ( t ) =1\), and \(q ( t ) =t^{-2}\). Then \(\phi ( t ) =\lambda ( t ) =t-t_{1}\), \(\vartheta ( t ) =\sum_{s=t_{2}} ^{t-1} ( s-t_{1} ) \), \(\gamma ( t ) =\beta ( t ) =t\). For \(k\in ( 0,1 ) \), it can be written \(kt \leq \phi ( t ) \leq t\), \(k^{2}t^{2}/2\leq \vartheta ( t ) \leq t^{2}/2\), \(k^{3}t^{3}/3\leq \sum_{s=t_{3}}^{t-1}k^{2}s ^{2}\leq t^{3}/3\). So,

$$\begin{aligned}& \lim_{t\rightarrow \infty }\sup \sum_{s=t_{3}}^{t-1} \Biggl( \frac{ \varGamma ( 1-\alpha ) \gamma ( s ) q ( s ) }{\vartheta ( s ) \phi ( s+1 ) }\sum_{u=t_{2}} ^{s-1} \frac{\vartheta ( u ) }{r ( u ) }\sum_{u=t _{1}}^{s-1} \frac{\phi ( u ) }{a ( u ) }-\frac{c ( s ) ( \Delta \gamma_{+} ( s ) ) ^{2}}{4 \gamma ( s ) } \Biggr) \\& \quad \geq \lim_{t\rightarrow \infty }\sup \sum_{s=t_{3}}^{t-1} \biggl( \frac{ \varGamma ( 1-\alpha ) k^{5}s^{2}}{6 ( s+1 ) }- \frac{1}{4s} \biggr) =\infty \end{aligned}$$

and

$$\begin{aligned}& \lim_{t\rightarrow \infty }\sup \sum_{\zeta =t_{2}}^{t-1} \Biggl( \frac{ \beta ( \zeta ) \lambda ( \zeta ) }{\lambda ( \zeta +1 ) a ( \zeta ) }\sum_{s=\zeta }^{\infty } \Biggl( \frac{1}{c ( s ) }\sum_{v=s}^{\infty }q ( v ) \Biggr) -\frac{r ( \zeta ) ( \Delta \beta_{+} ( \zeta ) ) ^{2}}{4\varGamma ( 1-\alpha ) \beta ( \zeta ) } \Biggr) \\& \quad \geq \lim_{t\rightarrow \infty }\sup \sum_{\zeta =t_{2}}^{t-1} \Biggl( \frac{\zeta^{2}}{ ( \zeta +1 ) }\sum_{s=\zeta }^{ \infty } \Biggl( \sum_{v=s}^{\infty }v^{-2} \Biggr) -\frac{1}{4\varGamma ( 1-\alpha ) \zeta } \Biggr) \\& \quad =\infty . \end{aligned}$$

Thus, (16) is oscillatory from Theorem 1.

References

  1. 1.

    Diethelm, K.: The Analysis of Fractional Differential Equations. Springer, Berlin (2010)

    Book  Google Scholar 

  2. 2.

    Kilbas, A., Srivastava, H., Trujillo, J.: Theory and Applications of Fractional Differential Equations. Elsevier, Amsterdam (2006)

    MATH  Google Scholar 

  3. 3.

    Öğrekçi, S.: Interval oscillation criteria for second-order functional differential equations. SIGMA 36(2), 351–359 (2018)

    Google Scholar 

  4. 4.

    Öğrekçi, S., Misir, A., Tiryaki, A.: On the oscillation of a second-order nonlinear differential equations with damping. Miskolc Math. Notes 18(1), 365–378 (2017)

    MathSciNet  Article  Google Scholar 

  5. 5.

    Öğrekçi, S.: New interval oscillation criteria for second-order functional differential equations with nonlinear damping. Open Math. 13(1), 239–246 (2015)

    MathSciNet  Article  Google Scholar 

  6. 6.

    Sadhasivam, V., Kavitha, J., Nagajothi, N.: Oscillation of neutral fractional order partial differential equations with damping term. Int. J. Pure Appl. Math. 115(9), 47–64 (2017)

    Google Scholar 

  7. 7.

    Hasil, P., Veselý, M.: Oscillation and non-oscillation criteria for linear and half-linear difference equations. J. Math. Anal. Appl. 452(1), 401–428 (2017)

    MathSciNet  Article  Google Scholar 

  8. 8.

    Hasil, P., Veselý, M.: Oscillation constants for half-linear difference equations with coefficients having mean values. Adv. Differ. Equ. 2015, 210 (2015)

    MathSciNet  Article  Google Scholar 

  9. 9.

    Sugie, J., Tanaka, M.: Nonoscillation theorems for second-order linear difference equations via the Riccati-type transformation. Proc. Am. Math. Soc. 145(5), 2059–2073 (2017)

    MathSciNet  Article  Google Scholar 

  10. 10.

    Zhang, C., Agarwal, R.P., Bohner, M., Li, T.: Oscillation of fourth-order delay dynamic equations. Sci. China Math. 58(1), 143–160 (2015)

    MathSciNet  Article  Google Scholar 

  11. 11.

    Grace, S.R., Agarwal, R.P., Sae-Jie, W.: Monotone and oscillatory behavior of certain fourth order nonlinear dynamic equations. Dyn. Syst. Appl. 19(1), 25–32 (2010)

    MathSciNet  MATH  Google Scholar 

  12. 12.

    Grace, S.R., Bohner, M., Sun, S.: Oscillation of fourth-order dynamic equations. Hacet. J. Math. Stat. 39, 545–553 (2010)

    MathSciNet  MATH  Google Scholar 

  13. 13.

    Grace, S.R., Argawal, R.P., Pinelas, S.: On the oscillation of fourth order superlinear dynamic equations on time scales. Dyn. Syst. Appl. 20, 45–54 (2011)

    MathSciNet  MATH  Google Scholar 

  14. 14.

    Li, T., Thandapani, E., Tang, S.: Oscillation theorems for fourth-order delay dynamic equations on time scales. Bull. Math. Anal. Appl. 3, 190–199 (2011)

    MathSciNet  MATH  Google Scholar 

  15. 15.

    Liu, T., Zheng, B., Meng, F.: Oscillation on a class of differential equations of fractional order. Math. Probl. Eng. 2013, Article ID 830836 (2013)

    MathSciNet  MATH  Google Scholar 

  16. 16.

    Qin, H., Zheng, B.: Oscillation of a class of fractional differential equations with damping term. Sci. World J. 2013, Article ID 685621 (2013)

    Google Scholar 

  17. 17.

    Ogrekci, S.: Interval oscillation criteria for functional differential equations of fractional order. Adv. Differ. Equ. 2015, 3 (2015)

    MathSciNet  Article  Google Scholar 

  18. 18.

    Muthulakshmi, V., Pavithra, S.: Interval oscillation criteria for forced fractional differential equations with mixed nonlinearities. Glob. J. Pure Appl. Math. 13(9), 6343–6353 (2017)

    Google Scholar 

  19. 19.

    Chen, D.-X.: Oscillation criteria of fractional differential equations. Adv. Differ. Equ. 2012, 33 (2012)

    MathSciNet  Article  Google Scholar 

  20. 20.

    Zheng, B.: Oscillation for a class of nonlinear fractional differential equations with damping term. J. Adv. Math. Stud. 6(1), 107–115 (2013)

    MathSciNet  MATH  Google Scholar 

  21. 21.

    Xu, R.: Oscillation criteria for nonlinear fractional differential equations. J. Appl. Math. 2013, Article ID 971357 (2013)

    MathSciNet  MATH  Google Scholar 

  22. 22.

    Secer, A., Adiguzel, H.: Oscillation of solutions for a class of nonlinear fractional difference equations. J. Nonlinear Sci. Appl. 9(11), 5862–5869 (2016)

    MathSciNet  Article  Google Scholar 

  23. 23.

    Abdalla, B.: On the oscillation of q-fractional difference equations. Adv. Differ. Equ. 2017, 254 (2017)

    MathSciNet  Article  Google Scholar 

  24. 24.

    Abdalla, B., Abodayeh, K., Abdeljawad, T. Alzabut, J.: New oscillation criteria for forced nonlinear fractional difference equations. Vietnam J. Math. 45(4), 609–618 (2017)

    MathSciNet  Article  Google Scholar 

  25. 25.

    Alzabut, J.O., Abdeljawad, T.: Sufficient conditions for the oscillation of nonlinear fractional difference equations. J. Fract. Calc. Appl. 5(1), 177–187 (2014)

    MathSciNet  Google Scholar 

  26. 26.

    Bai, Z., Xu, R.: The asymptotic behavior of solutions for a class of nonlinear fractional difference equations with damping term. Discrete Dyn. Nat. Soc. 2018, Article ID 5232147 (2018)

    MathSciNet  Article  Google Scholar 

  27. 27.

    Chatzarakis, G.E., Gokulraj, P., Kalaimani, T., Sadhasivam, V.: Oscillatory solutions of nonlinear fractional difference equations. Int. J. Differ. Equ. 13(1), 19–31 (2018)

    Google Scholar 

  28. 28.

    Sagayaraj, M.R., Selvam, A.G.M., Loganathan, M.P.: On the oscillation of nonlinear fractional difference equations. Math. Æterna 4, 220–224 (2014)

    MathSciNet  MATH  Google Scholar 

  29. 29.

    Selvam, A.G.M., Sagayaraj, M.R., Loganathan, M.P.: Oscillatory behavior of a class of fractional difference equations with damping. Int. J. Appl. Math. Res. 3(3), 220–224 (2014)

    Google Scholar 

  30. 30.

    Li, W.N.: Oscillation results for certain forced fractional difference equations with damping term. Adv. Differ. Equ. 2016, 70 (2016)

    MathSciNet  Article  Google Scholar 

  31. 31.

    Sagayaraj, M.R., Selvam, A.G.M., Loganathan, M.P.: Oscillation criteria for a class of discrete nonlinear fractional equations. Bull. Soc. Math. Serv. Stand. 3(1), 27–35 (2014)

    MATH  Google Scholar 

  32. 32.

    Atici, F.M., Eloe, P.W.: Initial value problems in discrete fractional calculus. Proc. Am. Math. Soc. 137(3), 981–989 (2008)

    MathSciNet  Article  Google Scholar 

  33. 33.

    Hardy, G.H., Littlewood, J.E., Polya, G.: Inequalities. Cambridge University Press, Cambridge (1988)

    MATH  Google Scholar 

Download references

Acknowledgements

The author is grateful to the scholars who provided the literature sources.

Availability of data and materials

Not applicable.

Funding

Not applicable.

Author information

Affiliations

Authors

Contributions

HA contributed to the work totally, and he read and approved the final version of the manuscript.

Corresponding author

Correspondence to Hakan Adiguzel.

Ethics declarations

Competing interests

The author declares that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Adiguzel, H. Oscillatory behavior of solutions of certain fractional difference equations. Adv Differ Equ 2018, 445 (2018). https://doi.org/10.1186/s13662-018-1905-3

Download citation

Keywords

  • Oscillation
  • Oscillation criteria
  • Fractional difference operator
  • Riemann–Liouville
  • Fractional difference equations
  • Riccati technique
  • Hardy inequalities