(ProQuest: ... denotes non-US-ASCII text omitted.)
Songbai Guo 1, 2 and Youjian Shen 2 and Binbin Shi 3
Academic Editor:Chuangxia Huang
1, School of Mathematics and Physics, University of Science and Technology Beijing, Beijing 100083, China
2, School of Mathematics and Statistics, Hainan Normal University, Haikou 571158, China
3, Department of Mathematics and Information Science, Shandong Agricultural University, Tai'an 271018, China
Received 31 December 2013; Accepted 5 April 2014; 27 April 2014
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.
1. Introduction
Delay differential equations (DDEs) arose widely in many fields, like oscillation theory [1-9], stability theory [10-12], dynamical behavior of delayed network systems [13-15], and so on. Theoretical studies on oscillation of solutions for DDEs have fundamental significance (see [16, 17]). For this reason, DDEs have been attracting great interest of many mathematicians during the last few decades.
In this paper, we consider a class of neutral DDEs [figure omitted; refer to PDF] where t 0 is a positive number and p , q , τ , and σ are positive constants. Generally, a solution of (1) is called oscillatory if it is neither eventually positive nor eventually negative. Otherwise, it is nonoscillatory. It can be seen in the literature that the oscillation theory regarding solutions of (1) has been extensively developed in the recent years.
In [18], Zhang came to the following conclusion.
Theorem I.
Assume that p ∈ ( 0,1 ) and q σ e > 1 - p ; then all solutions of (1) are oscillatory.
This result in Theorem I improves the corresponding result in [19]. Afterward, many authors have been devoted to studying this problem and have obtained many better results. For details, Gopalsamy and Zhang [20] obtained the improved result shown in Theorem II.
Theorem II.
If p ∈ ( 0,1 ) and q σ e > 1 - p [ 1 + q τ / ( 1 - p ) ] , then all solutions of (1) are oscillatory.
Further, Zhou and Yu [21] proved the following theorem.
Theorem III.
Suppose that p ∈ ( 0,1 ) and q σ e > 1 - p [ 1 + q τ / ( 1 - p ) + ( q τ ) 2 / 2 ( 1 - p ) 2 ] ; then all solutions of (1) are oscillatory.
Continuing to improve the research work, Xiao and Li [22] obtained the following.
Theorem IV.
Let p ∈ ( 0,1 ) and q σ e > 1 - p e q τ / ( 1 - p ) ; then all solutions of (1) are oscillatory.
Finally, Lin [23] obtained the result shown in Theorem V.
Theorem V.
Assume that p ∈ ( 0,1 ) and q σ e > 1 - p e q τ / ( 1 - p - q σ ) ; then all solutions of (1) are oscillatory.
However, all the conclusions mentioned above are limited to sufficient conditions in the case 0 < p < 1 . The aim of this paper is to establish systematically the necessary and sufficient conditions of oscillation for all solutions of (1) for the cases 0 < p < 1 and p > 1 .
2. Main Results
It is well known [24] that all solutions of (1) are oscillatory if and only if the characteristic equation of (1) [figure omitted; refer to PDF] has no real roots.
Theorem 1.
Assume that p ∈ ( 0,1 ) and let [figure omitted; refer to PDF] Then all solutions of (1) are oscillatory if and only if [figure omitted; refer to PDF] where θ is a unique zero of [straight phi] ( μ ) in ( 0 , 1 / σ ) .
Proof.
It is easy to see that, for λ [= or >, slanted] 0 , we have [figure omitted; refer to PDF] Thus any real root of (2) must be negative.
Next, let [figure omitted; refer to PDF] We consider the monotonicity of the function g ( μ ) ... = f ( - μ ) / μ . Differentiation yields [figure omitted; refer to PDF] where [straight phi] ( μ ) satisfies the following properties:
(1) [straight phi] ( μ ) > 0 for μ ∈ ( 1 / σ , + ∞ ) ;
(2) [straight phi] ( μ ) is strictly increasing on ( 0 , 1 / σ ) since the function μ 2 e ( τ - σ ) μ is strictly increasing on ( 0 , 1 / σ ) .
In addition, [figure omitted; refer to PDF] Thus, we get that function [straight phi] ( μ ) has a unique zero θ in ( 0 , 1 / σ ) . Hence g [variant prime] ( μ ) < 0 for μ ∈ ( 0 , θ ) and g [variant prime] ( μ ) > 0 for μ ∈ ( θ , + ∞ ) , which imply that g ( μ ) is decreasing on ( 0 , θ ) and increasing on ( θ , + ∞ ) . Therefore, g ( μ ) > 0 for μ ∈ ( 0 , + ∞ ) if and only if (7) has no real roots in μ ∈ ( 0 , 1 / σ ) . It is easy to see that g ( θ ) is the minimum value of g ( μ ) in ( 0 , 1 / σ ) . Consequently, g ( μ ) = 0 has no real roots in ( 0 , 1 / σ ) if and only if g ( θ ) > 0 . Since [figure omitted; refer to PDF] we obtain the result immediately.
From Theorem 1, we obtain immediately the following.
Corollary 2.
If p ∈ ( 0,1 ) and τ = σ , then all solutions of (1) are oscillatory if and only if q e θ σ > σ θ 2 holds, where θ = ( q σ ( q σ + 4 p ) - q σ ) / 2 p σ .
Theorem 3.
Suppose that p ∈ ( 0,1 ) ; then all solutions of (1) are oscillatory if and only if one of the following conditions holds:
( H 1 ): q σ e [= or >, slanted] 1 ;
( H 2 ): θ ¯ > θ ,
where θ and θ ¯ are the unique zeros of [straight phi] ( μ ) and h ( μ ) (see (3) and (4)) in ( 0 , 1 / σ ) , respectively.
Proof.
Let y ( μ ) = h ( μ ) / μ 2 = q e μ σ ( ( τ - σ ) / μ + 1 / μ 2 ) - τ ; then [figure omitted; refer to PDF] where z ( μ ) = ( τ - σ ) σ μ 2 + ( 2 σ - τ ) μ - 2 , which satisfies [figure omitted; refer to PDF] If τ [= or >, slanted] σ , we get obviously that z ( μ ) < 0 for all μ ∈ ( 0 , 1 / σ ] . If τ < σ , we also get z ( μ ) < 0 for all μ ∈ ( 0 , 1 / σ ] since z [variant prime] ( 1 / σ ) = τ > 0 . Thus, z ( μ ) < 0 for all μ ∈ ( 0 , 1 / σ ] . From this and (11) we get that y [variant prime] ( μ ) < 0 for all μ ∈ ( 0 , 1 / σ ] . Consequently, y ( μ ) is strictly decreasing on ( 0 , 1 / σ ] . Further, [figure omitted; refer to PDF] Therefore, if q σ e [= or >, slanted] 1 , we have y ( θ ) > 0 . Hence h ( θ ) > 0 . If q σ e < 1 , we have y ( 1 / σ ) < 0 . Hence, it is easy to find that both functions y ( μ ) and h ( μ ) have an equal and unique zero θ ¯ ∈ ( 0 , 1 / σ ) . Consequently, h ( θ ) > 0 is equivalent to θ ¯ > θ .
From Theorem 1, all solutions of (1) are oscillatory if and only if one of (H1 ) or (H2 ) holds.
Theorem 4.
Assume that p ∈ ( 0,1 ) ; then all solutions of (1) are oscillatory if one of the following conditions holds:
( H 1 ): q / θ + q σ [= or >, slanted] 1 - p ;
( H 2 ): q σ e [= or >, slanted] 1 - p e q τ / ( 1 - p - q σ ) ,
where θ is a unique zero of [straight phi] ( μ ) in ( 0 , 1 / σ ) .
Proof.
If q / θ + q σ [= or >, slanted] 1 - p , we have that [figure omitted; refer to PDF] From the proof of Theorem 1, all solutions of (1) are oscillatory.
If q σ e [= or >, slanted] 1 - p e q τ / ( 1 - p - q σ ) , we suppose furthermore that q / θ + q σ < 1 - p (otherwise, all solutions of (1) are oscillatory by the above conclusion); that is, θ > q / ( 1 - p - q σ ) . Since q σ e is a minimum value of the function ( q / μ ) e μ σ at μ = 1 / σ , we have that [figure omitted; refer to PDF] and the result follows.
So far, for p ∈ ( 0,1 ) we have discussed the necessary and sufficient conditions of oscillation for all solutions of (1). Our results have perfected the results in [23] (see Theorem 4). Next, we will discuss the behavior of oscillation of solutions of (1) in the case p > 1 .
Lemma 5.
Let p > 1 ; then all solutions of (1) are oscillatory if and only if the equation [figure omitted; refer to PDF] has no real roots in ( - ln ... p / τ , 0 ) .
Proof.
By (14), we know that g ( μ ) > 0 for μ ∈ ( 0 , ∞ ) . It is not difficult to see that e μ σ / μ is strictly decreasing on ( - ∞ , 0 ) while e μ τ is strictly increasing on ( - ∞ , 0 ) . Notice that p e μ τ - 1 = 0 at μ = - ln ... p / τ ; we find that [figure omitted; refer to PDF] Hence, f ( λ ) has no real roots which is equivalent to g ( μ ) that has no real roots in ( - ln ... p / τ , 0 ) .
Theorem 6.
Suppose that p > 1 and τ = σ ; then all solutions of (1) are oscillatory if and only if [figure omitted; refer to PDF] where θ = ( - q σ ( q σ + 4 p ) - q σ ) / 2 p σ .
Proof.
It is similar to the proof of Theorem 1; g ( θ ) is the maximum value of g ( μ ) for μ ∈ ( - ∞ , 0 ) . This and Lemma 5 imply the result.
Theorem 7.
Assume that p > 1 and τ < σ ; then all solutions of (1) are oscillatory if and only if [figure omitted; refer to PDF] where θ is a unique zero of (3) in ( - ∞ , 0 ) .
Proof.
Firstly, we prove that [straight phi] ( μ ) has a unique zero θ in ( - ∞ , 0 ) . In fact, [figure omitted; refer to PDF] It is easy to verify that [straight phi] [variant prime] ( μ ) is strictly increasing on ( - ∞ , 0 ) . In addition, [figure omitted; refer to PDF] Therefore, [straight phi] [variant prime] ( μ ) has a unique zero ω 0 in ( - ∞ , 0 ) . Hence, [straight phi] ( μ ) is strictly decreasing on ( - ∞ , ω 0 ) and strictly increasing on ( ω 0 , 0 ) , so that [straight phi] ( μ ) has a unique zero θ in ( - ∞ , 0 ) as [straight phi] ( 0 ) = - q < 0 and [straight phi] ( μ ) [arrow right] + ∞ ( μ [arrow right] - ∞ ) .
Now, from (8), it follows that g ( θ ) is the maximum value of g ( μ ) in ( - ∞ , 0 ) . By (10), we know that (19) is equivalent to g ( μ ) < 0 for μ ∈ ( - ∞ , 0 ) .
From Theorem 7, we obtain the following corollary that extends Theorem 1 in [25] for τ < σ .
Corollary 8.
If p > 1 , τ < σ , and τ q e - ( σ / τ ) ln ... p [= or >, slanted] τ ln ... 2 p / ( σ ln ... p + τ ) , then all solutions of (1) are oscillatory.
Proof.
The inequality τ q e - ( σ / τ ) ln ... p [= or >, slanted] τ ln ... 2 p / ( σ ln ... p + τ ) is equivalent to [straight phi] ( - ln ... p / τ ) [= or <, slanted] 0 . From the proof of Theorem 7, we get that θ [= or <, slanted] - ln ... p / τ . This and (17) imply g ( θ ) < 0 ; therefore, h ( θ ) < 0 .
Theorem 9.
Suppose that p > 1 and τ > σ ; then all solutions of (1) are oscillatory if and only if one of the following conditions holds:
( H 1 ): q σ e 2 - 2 [= or >, slanted] ( 2 2 - 2 ) p τ / ( τ - σ ) ;
( H 2 ): σ ( τ - σ ) ω 1 2 + ( 2 σ - τ ) ω 1 [= or <, slanted] 2 ;
( H 3 ): h ( θ 2 ) = q e θ 2 σ [ ( τ - σ ) θ 2 + 1 ] - τ θ 2 2 < 0 ,
where ω 1 is a unique zero of [straight phi] [variant prime] ( μ ) in ( - 2 / ( τ - σ ) , ( 2 - 2 ) / ( τ - σ ) ) and θ 2 is the maximum negative zero of [straight phi] ( μ ) .
Proof.
By Lemma 5, all solutions of (1) are oscillatory if and only if [figure omitted; refer to PDF] From (20), we have that [figure omitted; refer to PDF] and [straight phi] [variant prime] ( μ ) is strictly decreasing on ( - 2 / ( τ - σ ) , ( 2 - 2 ) / ( τ - σ ) ) and strictly increasing on ( ( 2 - 2 ) / ( τ - σ ) , 0 ) . Thus, [straight phi] [variant prime] ( ( 2 - 2 ) / ( τ - σ ) ) is the minimum value of [straight phi] [variant prime] ( μ ) in ( - 2 / ( τ - σ ) , 0 ) .
( 1 ) If [straight phi] [variant prime] ( ( 2 - 2 ) / ( τ - σ ) ) [= or >, slanted] 0 , which is the case of (H1 ), we have that [figure omitted; refer to PDF] Combining (23) and (24), we obtain that [figure omitted; refer to PDF] This means that g ( μ ) is strictly decreasing on ( - ∞ , 0 ) and, consequently, [figure omitted; refer to PDF]
( 2 ) If [straight phi] [variant prime] ( ( 2 - 2 ) / ( τ - σ ) ) < 0 , [straight phi] [variant prime] ( μ ) has a unique zero ω 1 in ( - 2 / ( τ - σ ) , ( 2 - 2 ) / ( τ - σ ) ) and a unique zero ω 2 in ( ( 2 - 2 ) / ( τ - σ ) , 0 ) since [straight phi] [variant prime] ( - 2 / ( τ - σ ) ) = q σ > 0 . Hence [straight phi] ( μ ) is strictly increasing on ( - ∞ , ω 1 ) , strictly decreasing on ( ω 1 , ω 2 ) , and strictly increasing on ( ω 2 , 0 ) . Consequently, [straight phi] ( ω 1 ) is the maximum value of [straight phi] ( μ ) in ( - ∞ , ω 2 ) . Now, it is easy to find that (22) holds if [straight phi] ( ω 1 ) [= or <, slanted] 0 .
On the other hand, applying [straight phi] [variant prime] ( ω 1 ) = 0 , we can get [figure omitted; refer to PDF] So [straight phi] ( ω 1 ) [= or <, slanted] 0 is equivalent to σ ( τ - σ ) ω 1 2 + ( 2 σ - τ ) ω 1 [= or <, slanted] 2 . This is the case of (H2 ).
If [straight phi] ( ω 1 ) > 0 , we obtain that [straight phi] ( μ ) has a unique zero θ 1 in ( - ∞ , ω 1 ) and a unique zero θ 2 in ( ω 1 , ω 2 ) . Therefore, g ( μ ) is strictly decreasing on ( - ∞ , θ 1 ) , strictly increasing on ( θ 1 , θ 2 ) , and strictly decreasing on ( θ 2 , 0 ) . Therefore, it is not difficult to find that (22) holds if and only if g ( θ 2 ) < 0 and it is the case of (H3 ).
From Theorem 9, we obtain the following corollary immediately.
Corollary 10.
If p > 1 , τ > σ , and q σ [= or >, slanted] p τ / ( τ - σ ) , then all solutions of (1) are oscillatory.
Example 11.
Consider the following neutral delay differential equation: [figure omitted; refer to PDF] It is not difficult to see that p = 20 , q = 10.5 , τ = 12 , and σ = 2 . Consequently, τ > σ , and [figure omitted; refer to PDF] so that all the solutions of (28) are oscillatory from Theorem 9.
Acknowledgments
The authors wish to thank the editor and anonymous reviewers for their helpful and valuable comments. This work was supported in part by NSF of Hainan Province under Grant 111004.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
[1] I. Györi, G. Ladas Oscillation Theory of Delay Differential Equations with Applications , of Oxford Mathematical Monographs, Clarendon Press, Oxford, UK, 1991.
[2] S. Tanaka, "Oscillation of solutions of first-order neutral differential equations," Hiroshima Mathematical Journal , vol. 32, no. 1, pp. 79-85, 2002.
[3] E. M. Elabbasy, T. S. Hassan, S. H. Saker, "Necessary and sufficient condition for oscillations of neutral differential equation," Serdica Mathematical Journal , vol. 31, no. 4, pp. 279-290, 2005.
[4] I. Kubiaczyk, S. H. Saker, J. Morchalo, "New oscillation criteria for first order nonlinear neutral delay differential equations," Applied Mathematics and Computation , vol. 142, no. 2-3, pp. 225-242, 2003.
[5] W.-T. Li, S. H. Saker, "Oscillation of nonlinear neutral delay differential equations with applications," Annales Polonici Mathematici , vol. 77, no. 1, pp. 39-51, 2001.
[6] G. Qin, C. Huang, Y. Xie, F. Wen, "Asymptotic behavior for third-order quasilinear differential equations," Advances in Difference Equations , vol. 2013, article 305, 2013.
[7] M. B. Dimitrova, V. I. Donev, "Sufficient conditions for oscillation of solutions of first order neutral delay impulsive differential equations with constant coefficients," Nonlinear Oscillations , vol. 13, no. 1, pp. 17-34, 2010.
[8] J. F. Gao, "Oscillations analysis of numerical solutions for neutral delay differential equations," International Journal of Computer Mathematics , vol. 88, no. 12, pp. 2648-2665, 2011.
[9] S. Guo, Y. Shen, "Necessary and sufficient conditions for oscillation of first order neutral difference equations," Acta Mathematicae Applicatae Sinica , vol. 36, no. 5, pp. 840-850, 2013.
[10] T. Naito, T. Hara, Y. Hino, R. Miyazaki, W. Ma, Z. Lu, "Differential equations with time lag," (Chinese) Introduction To Functional Differential Equations , Sciences Press, Beijing, China, 2013.
[11] G. Liu, J. Yan, "Global asymptotic stability of nonlinear neutral differential equation," Communications in Nonlinear Science and Numerical Simulation , vol. 19, no. 4, pp. 1035-1041, 2014.
[12] C. Huang, Z. Yang, T. Yi, X. Zou, "On the basins of attraction for a class of delay differential equations with non-monotone bistable nonlinearities," Journal of Differential Equations , vol. 256, no. 7, pp. 2101-2114, 2014.
[13] Q. Zhang, X. Wei, J. Xu, "Stability analysis for cellular neural networks with variable delays," Chaos, Solitons & Fractals , vol. 28, no. 2, pp. 331-336, 2006.
[14] G. He, J. Cao, "Discussion of periodic solutions for p th order delayed NDEs," Applied Mathematics and Computation , vol. 129, no. 2-3, pp. 391-405, 2002.
[15] C. Huang, H. Kuang, X. Chen, F. Wen, "An LMI approach for dynamics of switched cellular neural networks with mixed delays," Abstract and Applied Analysis , vol. 2013, 2013.
[16] J. Hale Theory of Functional Differential Equations , Springer, New York, NY, USA, 1977., 2nd.
[17] V. B. Kolmanvskii, V. R. Nosov Stability of Functional Differential Equations , Academic Process, New York, NY, USA, 1986.
[18] B. G. Zhang, "Oscillation of first order neutral functional-differential equations," Journal of Mathematical Analysis and Applications , vol. 139, no. 2, pp. 311-318, 1989.
[19] G. Ladas, Y. G. Sficas, "Oscillations of neutral delay differential equations," Canadian Mathematical Bulletin , vol. 29, no. 4, pp. 438-445, 1986.
[20] K. Gopalsamy, B. G. Zhang, "Oscillation and nonoscillation in first order neutral differential equations," Journal of Mathematical Analysis and Applications , vol. 151, no. 1, pp. 42-57, 1990.
[21] Y. Zhou, Y. H. Yu, "Oscillation for first order neutral delay differential equations," Journal of Mathematical Research and Exposition , vol. 21, no. 1, pp. 86-88, 2001.
[22] G. Xiao, X. Li, "A new oscillatory criterion for first order neutral delay differential equations," Journal of Nanhua University (Science & Engineering Edition) , vol. 15, no. 4, pp. 8-9, 2001.
[23] S. Lin, "Oscillation in first order neutral differential equations," Annals of Differential Equations , vol. 19, no. 3, pp. 334-336, 2003.
[24] G. Ladas, S. W. Schults, "On oscillations of neutral equations with mixed arguments," Hiroshima Mathematical Journal , vol. 19, no. 2, pp. 409-429, 1989.
[25] Y. Zhang, Y. Wang, "Oscillatory criteria for a class of first order neutral delay differential equations," Journal of Shanxi University (Natural Science Edition) , vol. 29, no. 4, pp. 341-342, 2006.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Copyright © 2014 Songbai Guo et al. Songbai Guo 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.
Abstract
We are concerned with oscillation of the first order neutral delay differential equation [superscript] [ x ( t ) - p x ( t - τ ) ] [variant prime] [/superscript] + q x ( t - σ ) = 0 with constant coefficients, and we obtain some necessary and sufficient conditions of oscillation for all the solutions in respective cases 0 < p < 1 and p > 1 .
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer