Download Full Issue in PDF - Academy Publisher
Download Full Issue in PDF - Academy Publisher
Download Full Issue in PDF - Academy Publisher
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
1530 JOURNAL OF COMPUTERS, VOL. 8, NO. 6, JUNE 2013<br />
The derivative and the forward jump operator are<br />
related by the useful formula<br />
σ<br />
f f μ f<br />
Δ<br />
σ<br />
= + , where f : = f σ.<br />
We will also make use of the follow<strong>in</strong>g product and<br />
quotient rules for the derivative of the product f g and the<br />
quotient f g( gg σ ≠ 0) of two differentiable functions f<br />
and g :<br />
Δ<br />
⎛ f ⎞ f g−<br />
f g<br />
( f g) Δ = f Δ g+ f σ<br />
g<br />
Δ<br />
, and ⎜ ⎟ =<br />
⎝ g ⎠ gg σ<br />
Δ<br />
Δ<br />
. (6)<br />
By us<strong>in</strong>g the product rule, the derivative of f () t =<br />
( t − α) m<br />
for m ∈ andα ∈ T can be calculated as<br />
m−1<br />
Δ<br />
v<br />
m−v−1<br />
() = ( σ() −α)( −α) .<br />
v=<br />
0<br />
f t ∑ t t<br />
(7)<br />
For a, b∈ T and a differentiable function f , the<br />
Cauchy <strong>in</strong>tegral of f Δ is def<strong>in</strong>ed by<br />
b Δ<br />
∫a f () t Δ t = f( b) − f( a)<br />
.<br />
The <strong>in</strong>tegration by parts formula follows from (6) and<br />
reads<br />
b Δ<br />
b b σ Δ<br />
∫ f () tgt () Δ t= ftgt () ()| −∫ f tg()<br />
tΔt.<br />
a<br />
To prove our ma<strong>in</strong> results, we will use the formula<br />
1 1<br />
( x γ Δ<br />
( t)) [ hx σ (1 h) x] γ − Δ<br />
= γ + − dhx ( t)<br />
0<br />
a<br />
a<br />
∫ . (8)<br />
which is a simple consequence of Keller′s cha<strong>in</strong> rule [2].<br />
Also, we need the follow<strong>in</strong>g lemma [5].<br />
Lemma 1 Assume A and B are nonnegtive constants, λ<br />
> 1, then<br />
λ<br />
−<br />
− ≤( − 1) .<br />
1<br />
AB λ A λ λ B<br />
λ<br />
The reader is referred to [2] for more detailed and<br />
extensive developments <strong>in</strong> calculus on time scales.<br />
III. MAIN RESULTS<br />
First, we state the oscillation criteria for (4).<br />
Set<br />
yt () = xt () + ctx () ( τ ()). t<br />
(9)<br />
Theorem 1 Assume that (H1) - (H5) hold, Furthermore,<br />
suppose that there exists a positive Δ−differentiable<br />
function g()<br />
t such that for all sufficiently large T ∗<br />
, and<br />
∗<br />
for all δ (T) > T , we have<br />
t<br />
∗<br />
limsup ∫ ( β (, sT ) gsQs () () −<br />
t→∞<br />
T<br />
α( s)(( g ( s)) )<br />
( γ + 1) g ( s)<br />
Δ γ + 1<br />
+<br />
γ+<br />
1 γ<br />
) Δ s =∞.<br />
Then every solution of (4) is oscillatory on [ t , ∞)<br />
0 T<br />
.<br />
(10)<br />
proof Suppose (4) has a nonoscillatory solution x (t).<br />
without loss of generality, there exists some t 1<br />
≥ t 0<br />
,<br />
sufficiently large such that xt () > 0, x( τ ( t)) > 0, x( δ ( t))<br />
> 0 for all t ≥ t . Hence In the view of (9), by (H1) we<br />
1<br />
get yt () > 0. from (4) and by (H2) - (H5), we have that<br />
Δ<br />
( y )) γ<br />
γ<br />
α<br />
Δ<br />
≤ −( q() t − p()) t x ( δ()) t < 0,<br />
and us<strong>in</strong>g the same proof of Theorem 1 [4], there exists<br />
t ≥ t such that for all t ≥ t , we have<br />
2 1<br />
2<br />
Δ<br />
⎧yt () > 0, y () t > 0,<br />
(11)<br />
⎨<br />
Δ<br />
( ( y ) γ Δ<br />
γ<br />
⎩ α ) ≤ −( q( t) − p( t))(1 − c( δ( t))) y ( δ( t)) < 0.<br />
By the def<strong>in</strong>ition of Qt (), we get<br />
Δ<br />
( ( y ) γ<br />
γ<br />
α )<br />
Δ<br />
≤ − Q( t) y ( δ( t)) < 0. (12)<br />
Make the generalized Riccati substitution<br />
Δ γ<br />
α()( t y ()) t<br />
wt () = gt ()<br />
. (13)<br />
γ<br />
y () t<br />
By the product and quotient rules, we have for all t ≥ t2<br />
Δ γ Δ<br />
Δ gt ()( α()( t y())) t ⎛ gt () ⎞<br />
Δ γ<br />
w () t = ( ()( t y ())) t<br />
γ<br />
+⎜ γ ⎟ α<br />
y () t ⎝ y () t ⎠<br />
Δ γ Δ<br />
gt ()( α()( t y()))<br />
t<br />
= +<br />
γ<br />
y () t<br />
Δ<br />
γ Δ<br />
g () t g()( t y ()) t<br />
Δ γ σ<br />
( −<br />
)( α( t)( y ( t)) ) .<br />
γσ γ γσ<br />
y () t y () t y () t<br />
From (12) - (14), we obta<strong>in</strong><br />
Δ<br />
Δ<br />
⎛ y( δ ( t)) ⎞ g ( t)<br />
σ<br />
w () t ≤− g() t Q() t ⎜ ⎟ + w () t<br />
σ<br />
⎝ yt () ⎠ g () t<br />
σ γ Δ<br />
gtw () ()( t y()) t<br />
−<br />
σ<br />
γ .<br />
g () t y () t<br />
γ<br />
Δ<br />
σ<br />
(14)<br />
(15)<br />
First consider the case when δ () t ≥ t . For all large t,<br />
Δ<br />
from y () t > 0, we have<br />
which implies that<br />
y( δ ( t))<br />
≥ 1 ,<br />
yt ()<br />
Δ<br />
σ γ Δ<br />
Δ<br />
g () t σ g() t w ()( t y ()) t<br />
w () t ≤− g() t Q() t + w () t − . (16)<br />
σ σ γ<br />
g () t g () t y () t<br />
Next consider the case when δ () t ≤ t, for all large t. By<br />
us<strong>in</strong>g α( y<br />
Δ )<br />
γ<br />
is strictly decreas<strong>in</strong>g on [ t 2<br />
, ∞ ) , we can<br />
choose t ≥ t such that δ () t ≥ t , for t ≥ t . Then we<br />
3 2<br />
2<br />
3<br />
obta<strong>in</strong><br />
© 2013 ACADEMY PUBLISHER