Hindawi Publishing Corporation Abstract and Applied Analysis Volume 2011, Article ID 693890, 23 pages doi:10.1155/2011/693890 Research Article Existence of Nonoscillatory Solutions for a Third-Order Nonlinear Neutral Delay Differential Equation Zeqing Liu,1 Lin Chen,1 Shin Min Kang,2 and Sun Young Cho3 1 Department of Mathematics, Liaoning Normal University, Dalian, Liaoning 116029, China Department of Mathematics and RINS, Gyeongsang National University, Jinju 660-701, Republic of Korea 3 Department of Mathematics, Gyeongsang National University, Jinju 660-701, Republic of Korea 2 Correspondence should be addressed to Shin Min Kang, smkang@gnu.ac.kr Received 4 April 2011; Accepted 28 June 2011 Academic Editor: Elena Braverman Copyright q 2011 Zeqing Liu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The aim of this paper is to study the solvability of a third-order nonlinear neutral delay differential equation of the form {αtβtxt ptxt − τ } ft, xσ1 t, xσ2 t, . . . , xσn t 0, t ≥ t0 . By using the Krasnoselskii’s fixed point theorem and the Schauder’s fixed point theorem, we demonstrate the existence of uncountably many bounded nonoscillatory solutions for the above differential equation. Several nontrivial examples are given to illustrate our results. 1. Introduction and Preliminaries In recent years, the study of the oscillation, nonoscillation, asymptotic behaviors and existence of solutions for various kinds of first- and second-order neutral delay differential equations and systems of differential equations have attracted much attention, for example, see 1–12 and the references therein. Dorociaková and Olach 2 discussed the existence of nonoscillatory solutions and asymptotic behaviors for the first-order delay differential equation x t ptxt qtxτt 0, t ≥ 0. 1.1 2 Abstract and Applied Analysis Elbert 3 and Huang 5 established a few oscillation and nonoscillation criteria for the second-order linear differential equation x t qtxt 0, t ≥ 0, 1.2 where q ∈ C0, ∞, R . Tang and Liu 10 studied the existence of bounded oscillation for the second-order linear delay differential equation of unstable type x t ptxt − τ, 1.3 t ≥ t0 , where τ > 0, p ∈ Ct0 , ∞, R and pt / ≡ 0 on any interval of length τ. In view of the Banach fixed point theorem, Kulenović and Hadžiomerspahić 7 deduced the existence of a nonoscillatory solution for the second-order linear neutral delay differential equation with positive and negative coefficients xt cxt − τ Q1 txt − σ1 − Q2 txt − σ2 0, t ≥ t0 , 1.4 where c ∈ R\{−1, 1}, τ > 0, σ1 , σ2 ∈ 0, ∞, Q1 , Q2 ∈ Ct0 , ∞, R . Qin et al. 9 and Yang et al. 11 developed several oscillation criteria for the second-order differential equation rt xt ptxt − τ qtfxt − δ 0, t ≥ t0 , 1.5 where τ and δ are nonnegative constants, r, p, q ∈ Ct0 , ∞, R, and f ∈ CR, R. By utilizing the Krasnoselskii’s fixed point theorem, Zhou 12 discussed the existence of nonoscillatory solutions of the second-order nonlinear neutral differential equation m rt xt ptxt − τ Qi tfxi t − σi 0, t ≥ t0 , 1.6 i1 where m ≥ 1 is an integer, τ > 0, σi ≥ 0, r, p, Qi ∈ Ct0 , ∞, R, and fi ∈ CR, R for i ∈ {1, 2, . . . , m}. However, the existence of nonoscillatory solutions of third-order neutral differential equations received much less attention, moreover, the results in 7, 11, 12 only figured out the existence of a nonoscillatory solution of 1.3–1.5, respectively. Motivated by the papers mentioned above, in this paper, we investigate the following third-order nonlinear neutral delay differential equation ft, xσ1 t, xσ2 t, . . . , xσn t 0, αt βt xt ptxt − τ t ≥ t0 , 1.7 where n ≥ 1 is an integer, τ > 0, α, β ∈ Ct0 , ∞, R \ {0}, p ∈ Ct0 , ∞, R, and f ∈ Ct0 , ∞ × Rn , R. By applying the Krasnoselskii’s fixed point theorem and the Schauder’s fixed point theorem, we obtain some sufficient conditions for the existence of uncountably many bounded nonoscillatory solutions of 1.7. Abstract and Applied Analysis 3 Throughout this paper, we assume that R −∞, ∞, R 0, ∞, Ct0 , ∞, R denotes the Banach space of all continuous and bounded functions on t0 , ∞ with the norm x supt≥t0 |xt| for each x ∈ Ct0 , ∞, R and AN, M {x ∈ Ct0 , ∞, R : N ≤ xt ≤ M, t ≥ t0 } for M > N > 0. 1.8 It is easy to see that AN, M is a bounded closed and convex subset of Ct0 , ∞, R. By a solution of 1.7, we mean a function x ∈ Ct1 − τ, ∞, R for some t1 ≥ t0 such that xt ptxt − τ, βtxt ptxt − τ and αtβtxt ptxt − τ are continuously differentiable in t1 , ∞ and such that 1.7 is satisfied for t ≥ t1 . As is customary, a solution of 1.7 is called oscillatory if it has arbitrarily large zeros, and otherwise it is said to be nonoscillatory. Definition 1.1 see 6. A family F of functions in Ct0 , ∞, R is equicontinuous on t0 , ∞ if for any ε > 0, the interval t0 , ∞ can be decomposed into a finite number of subintervals I1 , I2 , . . . , In such that fx − f y ≤ ε, ∀f ∈ F, x, y ∈ Ii , i ∈ {1, 2, . . . , n}. 1.9 Lemma 1.2 see Krasnoselskii’s fixed point theorem, 4. Let X be a Banach space. Let Ω be a bounded closed convex subset of X and S1 and S2 mappings from Ω into X such that S1 x S2 y ∈ Ω for every pair x, y ∈ Ω. If S1 is a contraction and S2 is completely continuous, then the equation S1 x S2 x x has at least one solution in Ω. Lemma 1.3 see Schauder’s fixed point theorem, 4. Let Ω be a nonempty closed convex subset of a Banach space X. Let S : Ω → Ω be a continuous mapping such that SΩ is a relatively compact subset of X. Then S has at least one fixed point in Ω. 2. Main Results Now we study those conditions under which 1.7 possesses uncountably many bounded nonoscillatory solutions. Theorem 2.1. Assume that there exist constants M, N, c1 , c2 , T0 and a function h ∈ Ct0 , ∞, R satisfying min{c1 , c2 } ≥ 0, c1 c2 < 1, 0 < N < 1 − c1 − c2 M, −c2 ≤ pt ≤ c1 , t ≥ T0 > t0 ; ft, u1 , u2 , . . . , un ≤ ht, t ∈ t0 , ∞, ui ∈ N, M, ∞ ∞ ∞ hu du ds dv < ∞. αsβv t0 v s 2.1 2.2 i ∈ {1, . . . , n}; Then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. 2.3 2.4 4 Abstract and Applied Analysis Proof. Set k ∈ c1 M N, 1 − c2 M. From 2.4, we pick up T > T0 such that ∞ ∞ ∞ T v s hu du ds dv < min{1 − c2 M − k, k − c1 M − N}. αsβv 2.5 Define two mappings S1k and S2k : AN, M → Ct0 , ∞, R by S1k xt S2k xt ⎧ ∞ ∞ ∞ ⎪ ⎨ ⎪ ⎩ t v s ⎧ ⎨k − ptxt − τ, t ≥ T, ⎩S xT , 1k t0 ≤ t < T, 2.6 1 fu, xσ1 u, . . . , xσn udu ds dv, αsβv t ≥ T, t0 ≤ t < T S2k xT , 2.7 for x ∈ AN, M. Firstly, we prove that S1k AN, M S2k AN, M ⊆ AN, M. By 2.1–2.3 and 2.5–2.7, we get that S1k xt S2k y t ≤ k c2 M ∞ ∞ ∞ T v s hu du ds dv αsβv ≤ k c2 M min{1 − c2 M − k, k − c1 M − N} ≤ M, x, y ∈ AN, M, t ≥ T, ∞ ∞ ∞ hu du ds dv S1k xt S2k y t ≥ k − c1 M − αsβv T v s 2.8 ≥ k − c1 M − min{1 − c2 M − k, k − c1 M − N} ≥ N, x, y ∈ AN, M, t ≥ T, which infer that S1k AN, M S2k AN, M ⊆ AN, M for any x, y ∈ AN, M. Secondly, we show that S1k is a contraction mapping. By 2.1, 2.2, and 2.6, we deduce that S1k xt − S1k y t ≤ ptxt − τ − yt − τ ≤ c1 c2 x − y 2.9 for x, y ∈ AN, M and t ≥ T , which gives that S1k x − S1k y ≤ c1 c2 x − y. 2.10 Abstract and Applied Analysis 5 Thirdly, we show that S2k is completely continuous. Let x ∈ AN, M and {xm }m≥1 ⊆ AN, M with limm → ∞ xm x. By 2.7, we obtain that |S2k xm t − S2k xt| ∞ ∞ ∞ 1 fu, xm σ1 u, . . . , xm σn u ≤ αsβv T v s −fu, xσ1 u, . . . , xσn udu ds dv, 2.11 t ≥ T, m ≥ 1. Using 2.3 and 2.4, we conclude that fu, xm σ1 u, . . . , xm σn u − fu, xσ1 u, . . . , xσn u ≤ 2hu, u ∈ T, ∞, m ≥ 1, ∞ fu, xm σ1 u, . . . , xm σn u − fu, xσ1 u, . . . , xσn udu ≤ 2 s ∞ hudu, s s ∈ T, ∞, m ≥ 1, ∞ ∞ v 1 fu, xm σ1 u, . . . , xm σn u − fu, xσ1 u, . . . , xσn udu ds αs s ∞ ∞ hu ≤2 du ds, v, s ∈ T, ∞, m ≥ 1. αs v s 2.12 In light of 2.11–2.12, fu, xm σ1 u, . . . , xm σn u − fu, xσ1 u, . . . , xσn u −→ 0 as m −→ ∞, 2.13 and the Lebesgue dominated convergence theorem, we conclude that lim S2k xm − S2k x 0, 2.14 m → ∞ which means that S2k is continuous in AN, M. Now we show that S2k is completely continuous. By virtue of 2.3, 2.4, and 2.7, we get that S2k x ≤ ∞ ∞ ∞ T v s hu du ds dv, αsβv x ∈ AN, M. 2.15 6 Abstract and Applied Analysis That is, S2k AN, M is uniform bounded. It follows from 2.4 that for each ε > 0, there exists T ∗ > T such that ∞ ∞ ∞ T∗ v s hu ε du ds dv < . αsβv 2 2.16 In view of 2.3, 2.7, and 2.16, we infer that |S2k xt2 − S2k xt1 | ∞ ∞ ∞ ∞ ∞ ∞ hu hu du ds dv du ds dv ≤ αsβv αsβv t2 v s t1 v s < ε, 2.17 x ∈ AN, M, t2 > t1 ≥ T ∗ . From 2.3 and 2.7, we get that |S2k xt2 − S2k xt1 | t2 ∞ ∞ hu ≤ du ds dv, αsβv s t1 v x ∈ AN, M, T ≤ t1 ≤ t2 ≤ T ∗ , 2.18 which together with 2.4 ensures that there exists δ > 0 satisfying |S2k xt2 − S2k xt1 | < ε, x ∈ AN, M, t1 , t2 ∈ T, T ∗ with |t2 − t1 | < δ. 2.19 Clearly, |S2k xt2 − S2k xt1 | 0 < ε, x ∈ AN, M, t0 ≤ t1 ≤ t2 ≤ T. 2.20 That is, S2k AN, M is equicontinuous on t0 , ∞. Consequently, S2k is completely continuous. By Lemma 1.2, there is x0 ∈ AN, M such that S1 x0 S2 x0 x0 , which is a bounded nonoscillatory solution of 1.7. Lastly, we demonstrate that 1.7 possesses uncountably many bounded nonoscilla k2 . For each j ∈ {1, 2}, tory solutions in AN, M. Let k1 , k2 ∈ c1 M N, 1 − c2 M and k1 / we choose Tj > T0 and the mappings Sjk1 and Sjk2 satisfying 2.5–2.7 with k and T replaced by kj and Tj , respectively, and some T3 > max{T1 , T2 } such that ∞ ∞ ∞ T3 v s hu |k1 − k2 | du ds dv < . αsβv 2 2.21 Obviously, there are x, y ∈ AN, M such that S1k1 x S2k1 x x and S1k2 y S2k2 y y, respectively. That is, x and y are two bounded nonoscillatory solutions of 1.7 in AN, M. Abstract and Applied Analysis 7 In order to prove that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M, we prove only that x / y. In fact, by 2.6 and 2.7, we gain that for t ≥ T3 xt k1 − ptxt − τ ∞ ∞ ∞ t yt k2 − ptyt − τ v s ∞ ∞ ∞ t v s 1 fu, xσ1 u, . . . , xσn udu ds dv, αsβv 1 f u, yσ1 u, . . . , yσn u du ds dv, αsβv 2.22 which together with 2.1–2.3 imply that xt − yt ≥ |k1 − k2 | − ptxt − τ − yt − τ − ∞ ∞ ∞ 1 αsβv t v s × fu, xσ1 u, . . . , xσn u − f u, yσ1 u, . . . , yσn u du ds dv ≥ |k1 − k2 | − c1 c2 x − y −2 ∞ ∞ ∞ T3 v s hu du ds dv, αsβv t ≥ T3 , 2.23 which means that x − y ≥ 1 1 c1 c2 |k1 − k2 | − 2 ∞ ∞ ∞ T3 v s hu du ds dv αsβv > 0, 2.24 by 2.1 and 2.21. That is, x / y. This completes the proof. Theorem 2.2. Assume that there exist constants M, N, c1 , c2 , T0 and a function h ∈ Ct0 , ∞, R satisfying 2.3, 2.4 and 0 < 1 − c2 N < 1 − c1 M; 0 ≤ c2 ≤ pt ≤ c1 < 1, t ≥ T 0 > t0 . 2.25 2.26 Then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Proof. Set k ∈ c1 M N, M c2 N. It follows from 2.4 that there exists T > T0 such that ∞ ∞ ∞ T v s hu du ds dv < min{M c2 N − k, k − c1 M − N}. αsβv 2.27 8 Abstract and Applied Analysis Define two mappings S1k and S2k : AN, M → Ct0 , ∞, R by 2.6 and 2.7, respectively. By virtue of 2.3, 2.6, 2.7, 2.26, and 2.27, we obtain that S1k xt S2k y t ≤ k − ptxt − τ ∞ ∞ ∞ T v s hu du ds dv αsβv ≤ k − c2 N min{M c2 N − k, k − c1 M − N} x, y ∈ AN, M, t ≥ T, ∞ ∞ ∞ S1k xt S2k y t ≥ k − ptxt − τ − 2.28 ≤ M, T v s hu du ds dv αsβv ≥ k − c1 M − min{M c2 N − k, k − c1 M − N} ≥ N, 2.29 x, y ∈ AN, M, t ≥ T, which yield that S1k AN, M S2k AN, M ⊆ AN, M. By a similar argument used in the proof of Theorem 2.1, we gain that S1k is a contraction mapping S2k is completely continuous, and 1.7 possesses uncountably many nonoscillatory solutions. This completes the proof. Theorem 2.3. Assume that there exist constants M, N, c1 , c2 , T0 and a function h ∈ Ct0 , ∞, R satisfying 2.3, 2.4, 2.25, and −1 < −c1 ≤ pt ≤ −c2 ≤ 0, t ≥ T 0 > t0 . 2.30 Then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Proof. Set k ∈ 1 − c2 N, 1 − c1 M. It follows from 2.4 that there exists T > T0 satisfying ∞ ∞ ∞ T v s hu du ds dv < min{1 − c1 M − k, k − 1 − c2 N}. αsβv 2.31 Let S1k and S2k : AN, M → Ct0 , ∞, R be defined by 2.6 and 2.7, respectively. In view of 2.3, 2.6, 2.7, 2.30, and 2.31, we obtain that S1 xt S2 y t ≤ k − ptxt − τ ∞ ∞ ∞ T v s hu du ds dv αsβv ≤ k c1 M min{1 − c1 M − k, k − 1 − c2 N} ≤ M, x, y ∈ AN, M, t ≥ T, ∞ ∞ ∞ S1 xt S2 y t ≥ k − ptxt − τ − T v s hu du ds dv αsβv ≥ k c2 N − min{1 − c1 M − k, k − 1 − c2 N} ≥ N, x, y ∈ AN, M, t ≥ T, which mean that S1k AN, M S2k AN, M ⊆ AN, M. 2.32 Abstract and Applied Analysis 9 The rest of the proof is similar to the proof of Theorem 2.1 and is omitted. This completes the proof. Theorem 2.4. Assume that there exist constants M and N with M > N > 0 and a function h ∈ Ct0 , ∞, R satisfying 2.3 and 2.4. If pt 1 for each t ∈ t0 , ∞, then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Proof. Set k ∈ N, M. It follows from 2.4 that there exists T ≥ t0 τ satisfying ∞ ∞ ∞ T v s hu du ds dv < min{M − k, k − N}. αsβv 2.33 Define a mapping Sk : AN, M → Ct0 , ∞, R by Sk xt ⎧ ∞ t2iτ ⎪ ⎨k ⎪ ⎩ i1 t2i−1τ ∞ ∞ v s 1 fu, xσ1 u, . . . , xσn udu ds dv, t ≥ T, αsβv t0 ≤ t < T 2.34 Sk xT , for x ∈ AN, M. Firstly, we prove that Sk AN, M ⊆ AN, M. By virtue of 2.3, 2.33 and 2.34, we obtain that Sk xt ≤ k ∞ t2iτ i1 ≤k Sk xt ≥ k − s t2i−1τ ∞ ∞ ∞ T ≥ N, v v s hu du ds dv αsβv hu du ds dv αsβv x ∈ AN, M, t ≥ T, ∞ ∞ ∞ t2iτ i1 ≥k− t2i−1τ ∞ ∞ ∞ T ≤ M, ∞ ∞ v s v s hu du ds dv αsβv hu du ds dv αsβv x ∈ AN, M, t ≥ T, which imply that Sk AN, M ⊆ AN, M. 2.35 10 Abstract and Applied Analysis Secondly, we show that Sk is continuous in AN, M. Let x ∈ AN, M and {xm }m≥1 ⊆ AN, M with limm → ∞ xm x. By 2.34, we get that |Sk xm t − Sk xt| ∞ ∞ ∞ t2iτ ≤ i1 ≤ t2i−1τ ∞ ∞ ∞ T v s v 1 fu, xm σ1 u, . . . , xm σn u αsβv −fu, xσ1 u, . . . , xσn udu ds dv s 1 fu, xm σ1 u, . . . , xm σn u αsβv −fu, xσ1 u, . . . , xσn udu ds dv, t ≥ T, m ≥ 1. 2.36 In view of 2.12, 2.13, 2.36, and the Lebesgue dominated convergence theorem, we deduce that lim Sk xm − Sk x 0, 2.37 m → ∞ which means that Sk is continuous in AN, M. Thirdly, we show that Sk AN, M is relatively compact. From 2.3, 2.33, and 2.34, we gain that Sk x ≤ k ∞ ∞ ∞ T v s hu du ds dv ≤ 2M, αsβv x ∈ AN, M, 2.38 which means that Sk AN, M is uniform bounded. Let ε > 0. It follows from 2.4 that there exist T > T ∗ > T such that ∞ ∞ ∞ T∗ v s ∞ ∞ ∞ v T s hu ε du ds dv < , αsβv 2 2.39 ε hu du ds dv < . αsβv 4 2.40 By 2.3, 2.34, and 2.39, we deduce that |Sk xt2 − Sk xt1 | ≤ t 2iτ ∞ 2 t2 2i−1τ i1 ≤2 ∞ ∞ ∞ < ε, T∗ v s t1 2iτ t1 2i−1τ ∞ ∞ v hu du ds dv αsβv x ∈ AN, M, t2 > t1 ≥ T ∗ . s hu du ds dv αsβv 2.41 Abstract and Applied Analysis 11 Choose an integer z ≥ 1 with T 2z 1τ ≥ T , and put ∞ ∞ hu du ds : v ∈ T τ, T ∗ 2z − 1τ , αsβv v s ∞ ∞ hu du ds : v ∈ T 2τ, T ∗ 2zτ , B max αsβv v s ε ε , . δ min 1 4Az 1 4Bz A max 2.42 Equation 2.4 means that max{A, B} < ∞. It follows from 2.3, 2.34, and 2.40–2.42 that for x ∈ AN, M, t1 , t2 ∈ T, T ∗ with t1 ≤ t2 ≤ t1 δ |Sk xt2 − Sk xt1 | t1 2iτ t2 2iτ z ≤ − i1 t2 2i−1τ t1 2i−1τ 1 fu, xσ1 u, . . . , xσn udu ds dv × αsβv v s t 2iτ t1 2iτ ∞ ∞ ∞ 2 hu du ds dv αsβv t2 2i−1τ t1 2i−1τ v s iz1 t2 2iτ t1 2i−1τ t1 2i−1τ z ≤ i1 t2 2i−1τ t1 2i−1τ t1 2iτ ∞ ∞ 1 fu, xσ1 u, . . . , xσn udu ds dv × αsβv v s ∞ ∞ 2 ∞ ∞ ∞ v T ≤ 2.43 z t 2i−1τ i1 t1 2i−1τ 2 s hu du ds dv αsβv t2 2iτ ∞ ∞ t1 2iτ ≤ Azt2 − t1 Bzt2 − t1 v s ε hu du ds dv αsβv 2 ε 2 < ε. It is not difficult to verify that |Sk xt2 − Sk xt1 | 0 < ε, x ∈ AN, M, t0 ≤ t1 ≤ t2 ≤ T. 2.44 12 Abstract and Applied Analysis Therefore Sk AN, M is equicontinuous on t0 , ∞, and consequently Sk is relatively compact. By Lemma 1.3, there is x0 ∈ AN, M such that Sx0 x0 , which together with 2.34 yields that for t ≥ T τ x0 t k ∞ t2iτ t2i−1τ i1 x0 t − τ k ∞ ∞ v s 1 fu, x0 σ1 u, . . . , x0 σn udu ds dv, αsβv ∞ t2i−1τ ∞ ∞ i1 t2i−1τ v s 2.45 1 fu, x0 σ1 u, . . . , x0 σn udu ds dv, αsβv which mean that x0 t x0 t − τ 2k ∞ ∞ ∞ t v s 1 fu, x0 σ1 u, . . . , x0 σn udu ds dv. 2.46 αsβv It is easy to show that x0 is a bounded nonoscillatory solution of 1.7. Finally, we demonstrate that 1.7 possesses uncountably many bounded nonoscilla k2 . For each j ∈ {1, 2}, we pick up a tory solutions in AN, M. Let k1 , k2 ∈ N, M and k1 / positive integer Tj ≥ t0 τ and the mapping Skj satisfying 2.33 and 2.34, where k and T are replaced by kj and Tj , respectively, and some T3 > max{T1 , T2 } satisfying 2.21. Clearly, there are x and y ∈ AN, M such that Sk1 x x and Sk2 y y, respectively. That is, x and y are bounded nonoscillatory solutions of 1.7 in AN, M. By 2.34 we get that for t ≥ T3 xt k1 ∞ t2iτ i1 yt k2 ∞ ∞ t2i−1τ ∞ t2iτ i1 v s ∞ ∞ t2i−1τ v s 1 fu, xσ1 u, . . . , xσn udu ds dv, αsβv 1 f u, yσ1 u, . . . , yσn u du ds dv, αsβv 2.47 which together with 2.3 and 2.21 yield that x − y ≥ |k1 − k2 | − 2 ∞ ∞ ∞ T3 v s hu du ds dv > 0, αsβv 2.48 which implies that x / y. This completes the proof. Theorem 2.5. Assume that there exist constants M and N with M > N > 0 and a function h ∈ Ct0 , ∞, R satisfying 2.3 and ∞ ∞ ∞ ∞ i1 t0 iτ v s hu du ds dv < ∞. αsβv 2.49 If pt −1 for each t ∈ t0 , ∞, then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Abstract and Applied Analysis 13 Proof. Set k ∈ N, M. It follows from 2.49 that there exists T > t0 satisfying ∞ ∞ ∞ ∞ T iτ i1 v hu du ds dv < min{M − k, k − N}. αsβv s 2.50 Define a mapping Sk : AN, M → Ct0 , ∞, R by Sk xt ⎧ ∞ ∞ ∞ ∞ ⎪ ⎨k − ⎪ ⎩ i1 tiτ v s 1 fu, xσ1 u, . . . , xσn udu ds dv, t ≥ T, αsβv t0 ≤ t < T 2.51 Sk xT , for x ∈ AN, M. Firstly, we prove that Sk AN, M ⊆ AN, M. By 2.3, 2.50, and 2.51, we obtain that Sk xt ≤ k ∞ ∞ ∞ ∞ tiτ i1 Sk xt ≥ k − v s ∞ ∞ ∞ ∞ tiτ i1 v s hu du ds dv ≤ M, αsβv x ∈ AN, M, t ≥ T, hu du ds dv ≥ N, αsβv x ∈ AN, M, t ≥ T, 2.52 which imply that SAN, M ⊆ AN, M. Secondly, we show that Sk is continuous in AN, M. Let x ∈ AN, M and {xm }m≥1 ⊆ AN, M with limm → ∞ xm x. By 2.51, we get that |Sk xm t − Sk xt| ∞ ∞ ∞ ∞ ≤ i1 T iτ v s 1 fu, xm σ1 u, . . . , xm σn u αsβv −fu, xσ1 u, . . . , xσn udu ds dv, t ≥ T, m ≥ 1. 2.53 Equation 2.53 together with 2.3, 2.49 and the Lebesgue dominated convergence theorem yields that lim Sk xm − Sk x 0, 2.54 m → ∞ that is, Sk is continuous in AN, M. Thirdly, we show that Sk AN, M is relatively compact. From 2.3, 2.50, and 2.51, we obtain that Sk x ≤ k ∞ ∞ ∞ ∞ i1 T iτ v s hu du ds dv ≤ 2M, αsβv x ∈ AN, M, 2.55 14 Abstract and Applied Analysis which means that Sk AN, M is uniform bounded. It follows from 2.49 that, for each ε > 0, there exists T ∗ > T such that ∞ ∞ ∞ ∞ T ∗ iτ i1 v s ε hu du ds dv < . αsβv 2 2.56 Notice that 2.3, 2.51, and 2.56 yield that ∞ ∞ ∞ ∞ |Sk xt2 − Sk xt1 | ≤ 2 i1 < ε, T ∗ iτ v s hu du ds dv αsβv 2.57 ∗ x ∈ AN, M, t2 > t1 ≥ T . Choose a positive integer z satisfying T zτ ≥ T ∗ . Put D max ∞ ∞ v s hu ∗ du ds : v ∈ T τ, T zτ , αsβv ε . δ 1 2Dz 2.58 By 2.3, 2.51, and 2.58, we gain that |Sk xt2 − Sk xt1 | ∞ ∞ ∞ ∞ ∞ 1 fu, xσ1 u, . . . , xσn udu ds dv − i1 αsβv t2 iτ t1 iτ v s ≤ ∞ t2 iτ ∞ ∞ i1 ≤ t1 iτ v s z t2 iτ ∞ ∞ i1 t1 iτ v ≤ Dzt2 − t1 s ∞ t2 iτ ∞ ∞ hu hu du ds dv du ds dv αsβv αsβv s iz1 t1 iτ v ∞ ∞ ∞ ∞ i1 < ε, hu du ds dv αsβv T ∗ iτ v s 2.59 hu du ds dv αsβv x ∈ AN, M, t1 , t2 ∈ T, T ∗ with t1 ≤ t2 ≤ t1 δ. Clearly, |Sk xt2 − Sk xt1 | 0 < ε, x ∈ AN, M, t0 ≤ t1 ≤ t2 ≤ T. 2.60 That is, Sk AN, M is equicontinuous on t0 , ∞, and Sk is relatively compact. The rest argument is similar to the proof of Theorem 2.4 and is omitted. This completes the proof. Abstract and Applied Analysis 15 Theorem 2.6. Assume that there exist constants M, N, c1 , c2 , T0 and a function h ∈ Ct0 , ∞, R satisfying 2.3, 2.4, and c12 − c2 c 2 − c1 N< 2 M; c1 c2 c2 ≤ pt ≤ c1 , t ≥ T0 ≥ t0 τ. 1 < c2 < c1 < c22 , 0< 2.61 2.62 Then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Proof. Set k ∈ c1 N c1 /c2 M, c2 M c2 /c1 N. It follows from 2.4 that there exists T > T0 such that ∞ ∞ ∞ T v s hu c2 c2 du ds dv < min c2 M N − k, k − M − c2 N . αsβv c1 c1 2.63 Define two mappings S1k and S2k : AN, M → Ct0 , ∞, R by ⎧ ⎪ ⎨ k xt τ − , pt τ pt τ S1k xt ⎪ ⎩S1k xT , S2k xt ⎧ ⎪ ⎨ ⎪ ⎩ 1 pt τ ∞ ∞ ∞ tτ v s t ≥ T, t0 ≤ t < T, 1 fu, xσ1 u, . . . , xσn udu ds dv, αsβv 2.64 t ≥ T, t0 ≤ t < T 2.65 S2k xT , for x ∈ AN, M. Firstly, we show that S1k AN, M S2k AN, M ⊆ AN, M. From 2.3, 2.61– 2.65, we get that S1k xt S2k y t ≤ k N 1 c2 c2 − min c2 M N − k, k − M − c2 N c2 c1 c2 c1 c1 ≤ M, x, y ∈ AN, M, t ≥ T, S1k xt S2k y t c2 c2 k M 1 − − min c2 M N − k, k − M − c2 N ≥ c1 c2 c2 c1 c1 ≥ N, x, y ∈ AN, M, t ≥ T, That is, S1k AN, M S2k AN, M ⊆ AN, M. 2.66 16 Abstract and Applied Analysis Secondly, by 2.61, 2.62, and 2.64, we conclude that S1k x − S1k y ≤ 1 x − y, c2 x, y ∈ AN, M, 2.67 which implies that S1k is a contraction mapping in AN, M. Thirdly, we show that S2k is completely continuous. Let x ∈ AN, M and {xm }m≥1 ⊆ AN, M be such that xm → x as m → ∞. By 2.61, 2.62, and 2.65, we gain that |S2k xm t − S2k xt| ∞ ∞ ∞ 1 1 fu, xm σ1 u, . . . , xm σn u ≤ pt τ tτ v αsβv s −fu, xσ1 t, . . . , xσn tdu ds dv 1 ∞ ∞ ∞ 1 fu, xm σ1 u, . . . , xm σn u ≤ c2 T αsβv v s −fu, xσ1 u, . . . , xσn udu ds dv, t ≥ T, m ≥ 1. 2.68 In view of 2.12, 2.13, 2.68, and the Lebesgue dominated convergence theorem, we obtain that lim S2k xm − S2k x 0, m → ∞ 2.69 that is, S2k is continuous in AN, M. For each ε > 0, it follows from 2.4 that there exists T ∗ ≥ T satisfying ∞ ∞ ∞ T∗ v s c2 ε c22 ε hu . du ds dv < min , αsβv 2 4c1 2.70 From 2.3, 2.61, 2.62, 2.65, and 2.70, we gain that |S2k xt2 − S2k xt1 | 1 ∞ ∞ ∞ hu 1 ∞ ∞ ∞ hu du ds dv du ds dv ≤ c2 t2 τ v αsβv c2 t1 τ v αsβv s s 2 ∞ ∞ ∞ hu du ds dv ≤ c2 T ∗ v αsβv s < ε, x ∈ AN, M, t2 > t1 ≥ T ∗ . 2.71 Abstract and Applied Analysis 17 By 2.3, 2.61, 2.62, 2.65, and 2.70, we obtain that, for x ∈ AN, M, T ≤ t1 ≤ t2 ≤ T ∗ , |S2k xt2 − S2k xt1 | t2 τ ∞ ∞ 1 hu ≤ du ds dv pt2 τ t1 τ v αsβv s ∞ ∞ ∞ 1 hu 1 − du ds dv pt2 τ pt1 τ t1 τ v αsβv s 1 t2 τ ∞ ∞ hu 2c1 ∞ ∞ ∞ hu ≤ du ds dv 2 du ds dv, c2 t1 τ v αsβv αsβv c2 T ∗ v s s 2.72 which means that there exists δ > 0 such that x ∈ AN, M, t1 , t2 ∈ T, T ∗ with |t2 − t1 | < δ. |S2k xt2 − S2k xt1 | < ε, 2.73 It is easy to verify that |S2k xt2 − S2k xt1 | 0 < ε, x ∈ AN, M, t0 ≤ t1 ≤ t2 ≤ T. 2.74 That is, S2k AN, M is equicontinuous on t0 , ∞, and S2k is completely continuous. By Lemma 1.2, there is x0 ∈ AN, M such that S1k x0 S2k x0 x0 , which is a bounded nonoscillatory solution of 1.7. Lastly, we demonstrate that 1.7 possesses uncountably many nonoscillatory k2 . For each j ∈ {1, 2}, solutions. Let k1 , k2 ∈ c1 N c1 /c2 M, c2 M c2 /c1 N and k1 / we choose Tj > t0 and contraction mappings Sjk1 and Sjk2 satisfying 2.63–2.65 with k and T replaced by kj and Tj , respectively, and some T3 > max{T1 , T2 } satisfying ∞ ∞ ∞ T3 v s c2 |k1 − k2 | hu du ds dv < . αsβv 2c1 2.75 Obviously, there are x, y ∈ AN, M such that S1k1 x S2k1 x x and S1k2 y S2k2 y y, respectively, and x and y are two bounded nonoscillatory solutions of 1.7 in AN, M. By 2.64 and 2.65, we gain that for t ≥ T3 xt xt τ k1 − pt τ pt τ ∞ ∞ ∞ 1 1 fu, xσ1 u, . . . , xσn udu ds dv, pt τ tτ v αsβv s yt τ k2 − yt pt τ pt τ ∞ ∞ ∞ 1 1 f u, yσ1 u, . . . , yσn u du ds dv, pt τ tτ v αsβv s 2.76 18 Abstract and Applied Analysis which together with 2.3, 2.61, and 2.62 mean that xt − yt ≥ ≥ 1 1 xt − τ − yt − τ |k1 − k2 | − pt τ pt τ ∞ ∞ ∞ 1 1 − pt τ tτ v αsβv s × fu, xσ1 u, . . . , xσn u − f u, yσ1 u, . . . , yσn u du ds dv 1 1 |k1 − k2 | − x − y c1 c2 ∞ ∞ ∞ 2 hu − du ds dv, c 2 T3 v αsβv s t ≥ T3 , 2.77 which together with 2.75 yields that x − y ≥ c2 1 c2 1 2 |k1 − k2 | − c1 c2 ∞ ∞ ∞ T3 v s hu du ds dv αsβv > 0, 2.78 that is, x / y. This completes the proof. Theorem 2.7. Assume that there exist constants M, N, c1 , c2 , T0 and a function h ∈ Ct0 , ∞, R satisfying 2.3, 2.4, and 0 < c1 − 1N < c2 − 1M; −∞ < −c1 ≤ pt ≤ −c2 < −1, t ≥ T0 ≥ t0 τ. 2.79 2.80 Then 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Proof. Set k ∈ c1 − 1N, c2 − 1M. It follows from 2.4 that there exists T > T0 such that ∞ ∞ ∞ T v s hu c2 k N du ds dv < min c2 − 1M − k, − c2 N . αsβv c1 2.81 Define two mappings S1k and S2k : AN, M → Ct0 , ∞, R by S1k xt S2k xt ⎧ ⎪ ⎨ ⎪ ⎩ 1 pt τ ∞ ∞ ∞ tτ S2k xT , v s ⎧ ⎪ ⎨ ⎪ ⎩ −k xt τ − , pt τ pt τ t ≥ T, S1k xT , t0 ≤ t < T, 1 fu, xσ1 u, . . . , xσn udu ds dv, αsβv t ≥ T, t0 ≤ t < T 2.82 Abstract and Applied Analysis 19 for x ∈ AN, M. From 2.3, 2.80–2.82, we obtain that S1k xt S2k y t k M 1 c2 k N min c2 − 1M − k, − c2 N c2 c2 c2 c1 ≤ ≤ M, x, y ∈ AN, M, t ≥ T, S1k xt S2k y t k N 1 c2 k N ≥ − min c2 − 1M − k, − c2 N , c1 c1 c2 c1 2.83 ≥ N, x, y ∈ AN, M, t ≥ T which implies that S1k AN, M S2k AN, M ⊆ AN, M. The rest of the proof is similar to the proof of Theorem 2.6 and is omitted. This completes the proof. 3. Examples In this section we construct seven examples as applications of the results presented in Section 2. Example 3.1. Consider the following third-order nonlinear neutral delay differential equation: ⎧ ⎨ t3 ⎩ et t ln1 t 1 et x2 ⎫ ⎬ 3 sin t2 − 2 1 xt − τ xt ⎭ 7 2 2t 1 3t x2 t − 1 t 3.1 0, t ≥ t0 , where τ > 0 and t0 > 0 are fixed. Let n 2, and let M and N be two positive constants with M > 7N and t3 αt t , e c1 t , βt ln1 t 4 , 7 c2 ft, u, v 2 , 7 σ1 t u2 v2 , et 3 sin t2 − 2 1 pt , 7 2t2 , 1 3t 1 σ2 t t − , t ht 2M2 , et 3.2 t, u, v ∈ t0 , ∞ × R2 . Obviously, 2.1–2.3 hold. On the other hand, ∞ ∞ ∞ t0 v s hu du ds dv M2 αsβv ∞ t0 ln1 v dv < ∞. v3 3.3 20 Abstract and Applied Analysis That is, 2.4 holds. Thus Theorem 2.1 means that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.2. Consider the following third-order nonlinear neutral delay differential equation: ⎧ ⎨ 2t 13 ⎩ t 2 4t 4t 5 xt 2t2 xt − τ 1 3t2 ⎫ ⎬ √ 2x t1 0, ⎭ t2 t ≥ t0 , 3.4 where τ > 0 and t0 > 0 are fixed. Let n 1, and let M and N be two positive constants with 3t20 1M > 3t20 1N and αt c1 2t 13 , t 2 , 3 ft, u c2 2u , t2 βt 4t2 4t 5, 2t20 , 2 1 3t0 σ1 t √ pt t 1, 2t2 , 1 3t2 ht 2M , t2 3.5 t, u ∈ t0 , ∞ × R. Obviously, 2.3, 2.25, and 2.26 hold. Moreover, ∞ ∞ ∞ t0 v s 2M 1 1π hu du ds dv − − arc tan2t0 1 < ∞. αsβv 16 22t0 1 2 2 3.6 Hence 2.4 holds. Thus Theorem 2.2 ensures that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.3. Consider the following third-order nonlinear neutral delay differential equation: e t−t2 1/3 −3t − 1 xt − τ t 1 xt 5t 2 4 tx t3 et2 1 x2 t2 2t 0, t ≥ 0, 3.7 where τ > 0 is fixed and t0 0. Let n 2, and let M and N be two positive constants with 4M > 5N and 2 αt et−t , c2 1 , 2 ft, u, v 1/3 βt t4 1 , σ1 t tt 2, tv , et2 1 u2 pt −3t − 1 , 5t 2 σ2 t t3 , t, u, v ∈ t0 , ∞ × R2 . ht c1 3 , 5 Mt , 1 N 2 et2 3.8 Abstract and Applied Analysis 21 Clearly, 2.3, 2.25, and 2.30 hold, and ∞ ∞ ∞ t0 v s M hu du ds dv αsβv 21 N 2 ∞ −1/3 e−v t4 1 dv < ∞, 3.9 t0 hence 2.4 holds. Therefore Theorem 2.3 guarantees that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.4. Consider the following third-order nonlinear neutral delay differential equation: x 2 t2 − t x t 2 − t 3 t3 1 1 0, xt xt − τ t2 arc tan t t t4 t ≥ t0 , 3.10 where τ > 0 and t0 > 0 are fixed. Let n 1, and let M and N be two positive constants with M > N and αt t3 1 , arc tan t t2 βt M2 M 3 , ht t4 1 , pt 1, t σ1 t t2 − t, u2 u 3 ft, u , t4 3.11 t, u ∈ t0 , ∞ × R. Obviously, 2.3 holds. Furthermore, ∞ ∞ ∞ t0 v s hu M2 M 3 du ds dv αsβv 3 ∞ t0 varc tan v dv < ∞, v3 1 3.12 that is, 2.4 holds. Thus Theorem 2.4 ensures that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.5. Consider the following third-order nonlinear neutral delay differential equation: 1 x2 t 1/t ln1 t 0, t3 xt − xt − τ t t3 t ≥ t0 , 3.13 where τ > 0 and t0 > 0 are fixed. Let n 1, and let M and N be two positive constants with M > N and αt t3 , βt M2 ht 3 , t 1 , t pt −1, u2 ft, u 3 , t σ1 t t 1 ln1 t, t t, u ∈ t0 , ∞ × R. 3.14 22 Abstract and Applied Analysis Obviously, 2.3 holds. Notice that ∞ ∞ ∞ ∞ i1 t0 iτ v s ∞ M2 hu 1 du ds dv < ∞, αsβv 16 i1 t0 iτ2 3.15 that is, 2.49 holds. Hence Theorem 2.5 implies that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.6. Consider the following third-order nonlinear neutral delay differential equation: t2 5π t 2 2/3 arc tan t t2 2 2 1 ln t xt xt − τ 0, 2 t 2 e e x t 1 1 t ≥ t0 , 3.16 where τ > 0 and t0 > 0 are fixed. Let n 1, and let M and N be two positive constants with 5/29π − 5/2N < 325π/4 − 3M and αt pt t2 , et2 βt 1 ln2 t, 5π arc tan t2/3 , 2 c1 3π, c2 5π , 2 t , 1et2 3.17 π hu 1 du ds dv − arc tanln t0 < ∞, 2 αsβv 2N 1 2 3.18 ft, u σ1 t t2 1, t , u2 1et2 ht N 2 t, u ∈ t0 , ∞ × R. Clearly, 2.3, 2.61, and 2.62 hold, and ∞ ∞ ∞ t0 v s that is, 2.4 holds also. Thus Theorem 2.6 ensures that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. Example 3.7. Consider the following third-order nonlinear neutral delay differential equation √ tx2 t2 1 2 1 t 1 1 xt sin t − 0, 4 cosln t − 7 xt − τ 2 t t t e et2 t ≥ t0 , 3.19 Abstract and Applied Analysis 23 where τ > 0 and t0 > 0 are fixed. Let n 1, and let M and N be two positive constants with M > 11N and t2 1 1 , βt , t et2 1 pt sin t − 4 cosln t − 7, t αt ht tM2 , et2 ft, u tu2 , et2 c1 12, c2 2, σ1 t ! t2 1, 3.20 t, u ∈ t0 , ∞ × R. Obviously, 2.3, 2.79, and 2.80 hold. Note that ∞ ∞ ∞ t0 v s M2 π hu du ds dv − arc tan t0 < ∞, αsβv 4 2 3.21 that is, 2.4 holds. It follows from Theorem 2.7 that 1.7 possesses uncountably many bounded nonoscillatory solutions in AN, M. References 1 J. Diblı́k and M. Růžičková, “Existence of positive solutions of a singular initial problem for a nonlinear system of differential equations,” The Rocky Mountain Journal of Mathematics, vol. 34, no. 3, pp. 923–944, 2004. 2 B. Dorociaková and R. Olach, “Existence of positive solutions of delay differential equations,” Tatra Mountains Mathematical Publications, vol. 43, pp. 63–70, 2009. 3 A. Elbert, “Oscillation/nonoscillation criteria for linear second order differential equations,” Journal of Mathematical Analysis and Applications, vol. 226, no. 1, pp. 207–219, 1998. 4 L. H. Erbe, Q. K. Kong, and B. G. Zhang, Oscillation Theory for Functional Differential Equations, Marcel Dekker, New York, NY, USA, 1995. 5 C. Huang, “Oscillation and nonoscillation for second order linear differential equations,” Journal of Mathematical Analysis and Applications, vol. 210, no. 2, pp. 712–723, 1997. 6 N. J. Huang and H. X. Li, “Existence of nonoscillatory solutions for nth order nonlinear neutral functional-differential equations,” Journal of Mathematical Analysis and Applications, vol. 216, no. 2, pp. 375–389, 1997. 7 M. R. S. Kulenović and S. Hadžiomerspahić, “Existence of nonoscillatory solution of second order linear neutral delay equation,” Journal of Mathematical Analysis and Applications, vol. 228, no. 2, pp. 436–448, 1998. 8 Z. Liu and S. M. Kang, “Infinitely many nonoscillatory solutions for second order nonlinear neutral delay differential equations,” Nonlinear Analysis, vol. 70, no. 12, pp. 4274–4293, 2009. 9 H. Z. Qin, N. N. Shang, and Y. M. Lu, “A note on oscillation criteria of second order nonlinear neutral delay differential equations,” Computers & Mathematics with Applications, vol. 56, no. 12, pp. 2987–2992, 2008. 10 X. H. Tang and Y. Liu, “Bounded oscillation for second-order delay differential equations with unstable type in a critical case,” Applied Mathematics Letters, vol. 16, no. 3, pp. 263–268, 2003. 11 Q. G. Yang, L. J. Yang, and S. M. Zhu, “Interval criteria for oscillation of second-order nonlinear neutral differential equations,” Computers & Mathematics with Applications, vol. 46, no. 5-6, pp. 903– 918, 2003. 12 Y. Zhou, “Existence for nonoscillatory solutions of second-order nonlinear differential equations,” Journal of Mathematical Analysis and Applications, vol. 331, no. 1, pp. 91–96, 2007. Advances in Operations Research Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Advances in Decision Sciences Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Mathematical Problems in Engineering Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Journal of Algebra Hindawi Publishing Corporation http://www.hindawi.com Probability and Statistics Volume 2014 The Scientific World Journal Hindawi Publishing Corporation http://www.hindawi.com Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 International Journal of Differential Equations Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Volume 2014 Submit your manuscripts at http://www.hindawi.com International Journal of Advances in Combinatorics Hindawi Publishing Corporation http://www.hindawi.com Mathematical Physics Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Journal of Complex Analysis Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 International Journal of Mathematics and Mathematical Sciences Journal of Hindawi Publishing Corporation http://www.hindawi.com Stochastic Analysis Abstract and Applied Analysis Hindawi Publishing Corporation http://www.hindawi.com Hindawi Publishing Corporation http://www.hindawi.com International Journal of Mathematics Volume 2014 Volume 2014 Discrete Dynamics in Nature and Society Volume 2014 Volume 2014 Journal of Journal of Discrete Mathematics Journal of Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Applied Mathematics Journal of Function Spaces Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Optimization Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014