 Research
 Open Access
Synchronization stability of delayed discretetime complex dynamical networks with randomly changing coupling strength
 MJ Park^{1},
 OM Kwon^{1}Email author,
 Ju H Park^{2},
 SM Lee^{3} and
 EJ Cha^{4}
https://doi.org/10.1186/168718472012208
© Park et al.; licensee Springer 2012
Received: 14 September 2012
Accepted: 19 November 2012
Published: 7 December 2012
Abstract
This paper addresses a delaydependent synchronization stability problem for discretetime complex dynamical networks with interval timevarying delays and randomly changing coupling strength. The randomly changing coupling strength is considered with the concept of binomial distribution. By constructing a suitable LyapunovKrasovskii functional and utilizing reciprocally convex approach and Finsler’s lemma, the proposed synchronization stability criteria for the networks are established in terms of linear matrix inequalities which can be easily solved by various effective optimization algorithms. The networks are represented by use of the Kronecker product technique. The effectiveness of the proposed methods will be verified via numerical examples.
Keywords
 Coupling Strength
 Bernoulli Random Variable
 Synchronization Criterion
 Binomial Random Variable
 Synchronization Analysis
1 Introduction
During the last few years, complex dynamic networks (CDNs), which is a set of interconnected nodes with specific dynamics, have received increasing attention from the real world such as the Internet, the World Wide Web, social networks, electrical power grids, global economic markets, and so on. Also, many models were proposed to describe various complex networks, smallworld network and scalefree network, etc. For more details, see the literature [1–4]. In the implementation of many practical CDNs, there exists timedelay because of the finite speed of information processing or amplifiers. It is well known that timedelay often causes undesirable dynamic behaviors such as oscillation and instability of the network. Therefore, various approaches to synchronization analysis for CDNs with timedelay have been investigated in the literature [5–12]. By using network modeling with coupling delays, Li et al. [5] proposed, for the first time, the synchronization criteria for the CDNs with timedelay which were expressed in the form of linear matrix inequalities (LMIs). Koo et al. [9] presented a synchronization criterion for singular CDNs with timevarying delays. In [10–12], various synchronization problems are addressed for discretetime CDNs with timedelay. In this regard, discretetime modeling with timedelay plays an important role in many fields of CDNs. Moreover, to implement the network, the network uses digital computers (usually a microprocessor or microcontrollers) with the necessary input/output hardware. The fundamental character of the digital computer is that it computes answers at discrete steps.
On the other hand, in [13–15], the problems for various systems with randomly occurring delay, uncertainties and nonlinearities were considered. The randomly occurring considerations in these literature works are described by the Bernoulli distribution. Here, the Bernoulli distribution is recognized as the experiment for the combination of U identical subexperiments. For more details, let A be the elementary event having one of the two possible outcomes as its element. $\overline{A}$ is the only other possible elementary event. At this time, we shall repeat the basic experiments U times and determine the probability that A is observed exactly v times out of U trials. Such repeated experiments are called Bernoulli trials [16]. Moveover, in [17], the stability of stochastic difference equations was addressed based on the Lyapunov functionals.
Regarding the CDNs, the coupling strength is the information of coupling strength between agents and a leader. Since an environmental change exists in practical networks, the change of coupling strength is a considerable factor affecting dynamics for the worse of the networks. At this point, the randomly changing coupling strength is being put to use in the problem of synchronization stability for CDNs. Moreover, to the best of authors’ knowledge, the synchronization analysis of CDNs with changing coupling strength has not been formulated yet.
Motivated by the results mentioned above, in this paper, a synchronization stability problem for discretetime CDNs with interval timevarying delays and randomly changing coupling strength will be studied. This information is one of randomly occurring coupling strength with binomial distribution. Put simply, the first and simplest random variable is the Bernoulli random variable. Let X be a random variable that takes on only two possible numerical values, $X(\mathrm{\Omega})=\{0,1\}$, where Ω represents the universal set consisting of the collection of all objects of interest in a particular context. Multiple independent Bernoulli random variables can be combined to construct more sophisticated random variables. Suppose X is the sum of w independent and identically distributed Bernoulli random variables. Then X is called a binomial random variable with parameters w, the number of trials, and p, the probability of success for each trial. Thus, the binomial distribution is a generalization of the Bernoulli distribution. Also, since delaydependent analysis makes use of the information on the size of time delay, more attention has been paid to the delaydependent analysis than to the delayindependent one [18]. That is, the former is generally less conservative than the latter. Therefore, a great number of results on a delaydependent stability condition for timedelay systems have been reported in the literature [19–24]. So, by construction of a suitable LyapunovKrasovskii functional and utilization of a reciprocally convex approach [24], a synchronization stability problem for discretetime CDNs with interval timevarying delays and randomly changing coupling strength is derived in terms of LMIs which can be solved efficiently by use of standard convex optimization algorithms such as interiorpoint methods [25]. Moreover, the discretetime CDNs are represented by use of the Kronecker product technique. Two numerical examples are included to show the effectiveness of the proposed methods.
Notation ${\mathbb{R}}^{n}$ is the ndimensional Euclidean space, and ${\mathbb{R}}^{m\times n}$ denotes the set of all $m\times n$ real matrices. For real symmetric matrices X and Y, $X>Y$ (respectively, $X\ge Y$) means that the matrix $XY$ is positive definite (respectively, nonnegative). ${X}^{\perp}$ denotes the basis for the nullspace of X. ${I}_{n}$, ${0}_{n}$, and ${0}_{m\times n}$ denote $n\times n$ identity matrix, $n\times n$ and $m\times n$ zero matrices, respectively. $\mathbb{E}\{\cdot \}$ stands for the mathematical expectation operator. $\parallel \cdot \parallel $ refers to the Euclidean vector norm or the induced matrix norm. $diag\{\cdots \}$ denotes the block diagonal matrix. ⋆ represents the elements below the main diagonal of a symmetric matrix. ${X}_{[f(t)]}\in {\mathbb{R}}^{m\times n}$ means that the elements of matrix ${X}_{[f(t)]}$ include the scalar value of $f(t)$.
2 Problem statements
Here, N is the number of coupled nodes, n is the number of state of each node, ${y}_{i}(k)={[{y}_{i1}(k),{y}_{i2}(k),\dots ,{y}_{in}(k)]}^{T}\in {\mathbb{R}}^{n}$ is the state vector of the i th node. $f({y}_{i}(k))=[f({y}_{i1}(k)),f({y}_{i2}(k)),\dots ,f({y}_{in}(k))]\in {\mathbb{R}}^{n}$ is a continuous differentiable vector function. The constant $c>0$ is the coupling strength. $h(k)$ is an interval timevarying delay satisfying $0\le {h}_{m}\le h(k)\le {h}_{M}$, where ${h}_{m}$ and ${h}_{M}$ are known positive integers.
In order to investigate the synchronization stability analysis for discretetime CDNs with interval timevarying delays in the coupling term (1), we introduce the following definition and lemmas.
Definition 1 ([5])
where $s(k)\in {\mathbb{R}}^{n}$ is a solution of an isolated node, satisfying $s(k+1)=f(s(k),s(kh(k)))$.
Lemma 1 ([11])
where J and ${J}_{d}$ are the Jacobian of $f(x(k),x(kh(k)))$ at $s(k)$ and $s(kh(k))$, respectively. Then the synchronized states (3) are asymptotically stable.
For the convenience of synchronization analysis for the system (4), the following Kronecker product and its properties are used.
Lemma 2 (Kronecker product [26])
 (i)
$(\alpha A)\otimes B=A\otimes (\alpha B)$,
 (ii)
$(A+B)\otimes C=A\otimes C+B\otimes C$,
 (iii)
$(A\otimes B)(C\otimes D)=(AC)\otimes (BD)$.
where N is the number of agents.
Moreover, it is assumed that the coupling strength has changed by the following assumption.
Remark 1 As mentioned in Section 1, the Bernoulli random variable takes on only two possible numerical values, $X(\mathrm{\Omega})=\{0,1\}$, where Ω represents the universal set consisting of the collection of all objects of interest in a particular context. However, with parameters w, the number of trials, and p, the probability of success for each trial, the binomial random variable X is the sum of w independent and identically distributed Bernoulli random variables. Thus, the former is more general than the latter. So, in this paper, the randomly changing coupling strength is considered with the concept of binomial distribution. In addition, the Bernoulli random variable has been used in the concept of randomly occurring which has various types such as randomly occurring delay, randomly occurring uncertainties, randomly occurring nonlinearities, and so on [13–15].
Remark 2 From Assumption 1, the coupling strength per discrete step is changed with the multiple of given strength, c, and the number of changes, m. Also, the probability of change is given as ${\rho}_{0}$. Based on the results mentioned above, for the concerned discretetime CDNs (1), the coupling strength with the binomial random variable can be represented in the model of discretetime CDNs with randomly changing coupling strength (6). It should be noted that this problem for the change of coupling strength has not been investigated yet.
The aim of this paper is to investigate the delaydependent synchronization stability analysis for the system (6). In order to do this, we introduce the following definition and lemmas.
Therefore, the inequality (9) is equivalent to the inequality (7) according to the Schur complement [25]. □
Lemma 4 (Finsler’s lemma [28])
 (i)
${\zeta}^{T}\mathrm{\Phi}\zeta <0$, $\mathrm{\forall}\mathrm{\Upsilon}\zeta =0$, $\zeta \ne 0$,
 (ii)
${{\mathrm{\Upsilon}}^{\perp}}^{T}\mathrm{\Phi}{\mathrm{\Upsilon}}^{\perp}<0$,
 (iii)
$\mathrm{\exists}F\in {\mathbb{R}}^{n\times m}:\mathrm{\Phi}+F\mathrm{\Upsilon}+{(F\mathrm{\Upsilon})}^{T}<0$.
3 Main results
Now, the following theorem is given for synchronization stability of the model of discretetime CDNs with interval timevarying delays in the coupling term (6).
where Φ and ${\mathrm{\Upsilon}}_{[{\rho}_{m}]}$ are defined in (10).
where $\alpha (k)=({h}_{M}h(k))/({h}_{M}{h}_{m})$.
respectively.
where ${\mathrm{\Upsilon}}_{[{\rho}_{m}]}$ is defined in (10).
where F is any matrix with appropriate dimension.
From the inequality (31), if the LMIs (11)(13) are satisfied, then the synchronization stability condition (29) holds by Definition 1. This completes our proof. □
The following is given for synchronization stability of the model of discretetime CDNs with only interval timevarying delays in nodes (33).
The other procedure is straightforward from the proof of Theorem 1, so it is omitted. □
Remark 3 The systems (6) and (33) with randomly changing coupling strength and the switched systems [30–38] are similar in the concept of changing parameters. In [30–38], the various problems for the switched neural networks with timeinvariant delay were addressed. However, since time delay has not only a fixed value in a practical system [39], the concerned systems with interval timevarying delays were considered in this paper. Moreover, the changing information of a parameter was considered with the probabilistic rule; that is, the Bernoulli sequence.
4 Numerical examples
In this section, we provide two numerical examples to show the effectiveness of the presented stability criteria in this paper.

Case 1:$G=\left[\begin{array}{ccccc}2& 1& 0& 0& 1\\ 1& 3& 1& 1& 0\\ 0& 1& 2& 1& 0\\ 0& 1& 1& 3& 1\\ 1& 0& 0& 1& 2\end{array}\right],$

Case 2:$G=\left[\begin{array}{cccccccccc}4& 1& 0& 0& 1& 1& 1& 0& 0& 0\\ 1& 3& 0& 0& 0& 0& 0& 1& 1& 0\\ 0& 0& 1& 0& 0& 0& 0& 1& 0& 0\\ 0& 0& 0& 2& 0& 1& 0& 1& 0& 0\\ 1& 0& 0& 0& 2& 0& 0& 0& 1& 0\\ 1& 0& 0& 1& 0& 2& 0& 0& 0& 0\\ 1& 0& 0& 0& 0& 0& 3& 0& 1& 1\\ 0& 1& 1& 1& 0& 0& 0& 3& 0& 0\\ 0& 1& 0& 0& 1& 0& 1& 0& 4& 1\\ 0& 0& 0& 0& 0& 0& 1& 0& 1& 2\end{array}\right].$
where the associated parameters are defined in Example 1.
5 Conclusions
In this paper, new delaydependent synchronization criteria for the discretetime CDNs with interval timevarying delays and randomly changing coupling strength are proposed. The randomly changing coupling strength is considered with the concept of binomial distribution, which is a generalization of the Bernoulli distribution. To drive these results, the suitable LyapunovKrasovskii functional and reciprocally convex approach are used to obtain the feasible region of synchronization stability criteria. Two numerical examples have been given to show the effectiveness and usefulness of the presented criteria.
Declarations
Acknowledgements
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (20120000479), and by a grant of the Korea Healthcare Technology R & D Project, Ministry of Health & Welfare, Republic of Korea (A100054).
Authors’ Affiliations
References
 Watts DJ, Strogatz SH: Collective dynamics of ‘smallworld’ networks. Nature 1998, 393: 440–442. 10.1038/30918View ArticleGoogle Scholar
 Strogatz SH: Exploring complex networks. Nature 2001, 410: 268–276. 10.1038/35065725View ArticleGoogle Scholar
 Li C, Xu C, Sun W, Xu J, Kurths J: Outer synchronization of coupled discretetime networks. Chaos 2009., 19: Article ID 013106Google Scholar
 Boccaletti S, Latora V, Moreno Y, Chavez M, Hwang DU: Complex networks: structure and dynamics. Phys. Rep. 2006, 424: 175–308. 10.1016/j.physrep.2005.10.009MathSciNetView ArticleGoogle Scholar
 Li C, Chen G: Synchronization in general complex dynamical networks with coupling delays. Physica A 2004, 343: 263–278.MathSciNetView ArticleGoogle Scholar
 Li K, Gong X, Guan S, Lai CH: Synchronization stability of general complex dynamical networks with timevarying delays. Phys. Lett. A 2008, 372: 7133–7139. 10.1016/j.physleta.2008.10.054MathSciNetView ArticleMATHGoogle Scholar
 Wang B, Guan ZH: Chaos synchronization in general complex dynamical networks with coupling delays. Nonlinear Anal., Real World Appl. 2010, 11: 1925–1932. 10.1016/j.nonrwa.2009.04.020MathSciNetView ArticleMATHGoogle Scholar
 Park MJ, Kwon OM, Park JH, Lee SM, Cha EJ: Synchronization criteria for coupled stochastic neural networks with timevarying delays and leakage delay. J. Franklin Inst. 2012, 349: 1699–1720. 10.1016/j.jfranklin.2012.02.002MathSciNetView ArticleMATHGoogle Scholar
 Koo JH, Ji DH, Won SC: Synchronization of singular complex dynamical networks with timevarying delays. Appl. Math. Comput. 2010, 217: 3916–3923. 10.1016/j.amc.2010.09.055MathSciNetView ArticleMATHGoogle Scholar
 Gao H, Lam J, Chen G: New criteria for synchronization stability of general complex dynamical networks with coupling delays. Phys. Lett. A 2006, 360: 263–273. 10.1016/j.physleta.2006.08.033View ArticleMATHGoogle Scholar
 Yue D, Li H: Synchronization stability of continuous/discrete complex dynamical networks with interval timevarying delays. Neurocomputing 2010, 73: 809–819. 10.1016/j.neucom.2009.10.008View ArticleGoogle Scholar
 Liu Y, Wang Z, Liu X, Liang J: Synchronization and state estimation for discretetime complex networks with distributed delays. IEEE Trans. Syst. Man Cybern., Part B, Cybern. 2008, 38: 1314–1325.View ArticleGoogle Scholar
 Bao H, Cao J: Synchronization of discretetime stochastic neural networks with random delay. Discrete Dyn. Nat. Soc. 2011. doi:10.1155/2011/713502Google Scholar
 Wu ZG, Park JH, Su H, Chu J: Robust dissipativity analysis of neural networks with timevarying delay and randomly occurring uncertainties. Nonlinear Dyn. 2012, 69: 1323–1332. 10.1007/s1107101203501MathSciNetView ArticleMATHGoogle Scholar
 Hu J, Wang Z, Gao H, Stergioulas LK: Robust sliding mode control for discrete stochastic systems with mixed time delays, randomly occurring uncertainties, and randomly occurring nonlinearities. IEEE Trans. Ind. Electron. 2012, 59: 3008–3015.View ArticleGoogle Scholar
 Peebels PZ: Probability, Random, Variables, and Random Signal Principles. McGrawHill, New York; 2001.Google Scholar
 Shaikhet L: Lyapunov Functionals and Stability of Stochastic Difference Equations. Springer, New York; 2011.View ArticleMATHGoogle Scholar
 Xu S, Lam J: A survey of linear matrix inequality techniques in stability analysis of delay systems. Int. J. Syst. Sci. 2008, 39: 1095–1113. 10.1080/00207720802300370MathSciNetView ArticleMATHGoogle Scholar
 Niculescu SI: Delay Effects on Stability: A Robust Approach. Springer, New York; 2002.Google Scholar
 Richard JP: Timedelay systems: an overview of some recent advances and open problems. Automatica 2003, 39: 1667–1694. 10.1016/S00051098(03)001675MathSciNetView ArticleMATHGoogle Scholar
 Gao HJ, Chen TW: New results on stability of discretetime systems with timevarying state delay. IEEE Trans. Autom. Control 2007, 52: 328–334.MathSciNetView ArticleGoogle Scholar
 Ahn CK: Fuzzy delayed output feedback synchronization for timedelayed chaotic systems. Nonlinear Anal. Hybrid Syst. 2010, 4: 16–24. 10.1016/j.nahs.2009.07.002MathSciNetView ArticleMATHGoogle Scholar
 Kwon OM, Park JH, Lee SM: An improved delaydependent criterion for asymptotic stability of uncertain dynamic systems with timevarying delays. J. Optim. Theory Appl. 2010, 145: 343–353. 10.1007/s109570099637xMathSciNetView ArticleMATHGoogle Scholar
 Park P, Ko JW, Jeong CK: Reciprocally convex approach to stability of systems with timevarying delays. Automatica 2011, 47: 235–238. 10.1016/j.automatica.2010.10.014MathSciNetView ArticleMATHGoogle Scholar
 Boyd S, Ghaoui LE, Feron E, Balakrishnan V: Linear Matrix Inequalities in System and Control Theory. SIAM, Philadelphia; 1994.View ArticleMATHGoogle Scholar
 Graham A: Kronecker Products and Matrix Calculus: With Applications. Wiley, New York; 1982.MATHGoogle Scholar
 Gu K: An integral inequality in the stability problem of timedelay systems. The 39th IEEE Conf. Decision Control 2000, 2805–2810.Google Scholar
 de Oliveira MC, Skelton RE: Stability tests for constrained linear systems. In Perspectives in Robust Control. Springer, Berlin; 2001:241–257.View ArticleGoogle Scholar
 Kim SH:Improved approach to robust ${H}_{\mathrm{\infty}}$ stabilization of discretetime TS fuzzy systems with timevarying delays. IEEE Trans. Fuzzy Syst. 2010, 18: 1008–1015.View ArticleGoogle Scholar
 Ahn CK: Sets of generalized H2 exponential stability criteria for switched multilayer dynamic neural networks. Adv. Differ. Equ. 2012. doi:10.1186/1687–1847–2012–150Google Scholar
 Ahn CK: Linear matrix inequality optimization approach to exponential robust filtering for switched Hopfield neural networks. J. Optim. Theory Appl. 2012, 154: 573–587. 10.1007/s1095701200087MathSciNetView ArticleMATHGoogle Scholar
 Ahn CK: An error passivation approach to filtering for switched neural networks with noise disturbance. Neural Comput. Appl. 2012, 21: 853–861. 10.1007/s0052101004745View ArticleGoogle Scholar
 Ahn CK: Switched exponential state estimation of neural networks based on passivity theory. Nonlinear Dyn. 2012, 67: 573–586. 10.1007/s110710110010xView ArticleMathSciNetMATHGoogle Scholar
 Ahn CK: Exponentially convergent state estimation for delayed switched recurrent neural networks. Eur. Phys. J. E 2011, 34: 122.View ArticleGoogle Scholar
 Ahn CK:${L}_{2}{L}_{\mathrm{\infty}}$ filtering for timedelayed switched Hopfield neural networks. Int. J. Innov. Comput. Inf. Control 2011, 7: 1831–1844.Google Scholar
 Ahn CK:Switched synchronization with a guaranteed${H}_{\mathrm{\infty}}$ performance. Chin. Phys. Lett. 2011., 28: Article ID 010501Google Scholar
 Ahn CK: Passive learning and inputtostate stability of switched Hopfield neural networks with timedelay. Inf. Sci. 2010, 180: 4582–4594. 10.1016/j.ins.2010.08.014View ArticleMATHGoogle Scholar
 Ahn CK:An ${H}_{\mathrm{\infty}}$ approach to stability analysis of switched Hopfield neural networks with timedelay. Nonlinear Dyn. 2010, 60: 703–711. 10.1007/s1107100996256View ArticleMathSciNetMATHGoogle Scholar
 Jiang L, Yao W, Wu QH, Wen JY, Cheng SJ: Delaydependent stability for load frequency control with constant and timevarying delays. IEEE Trans. Power Syst. 2012, 27: 932–941.View ArticleGoogle Scholar
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.