Communications in Nonlinear Science and Numerical Simulation
Network coherence in the web graphs
Introduction
Complex networks have been proved as an important tool for revealing the structures and functions of natural and man-made systems with special focus on some attributes, such as small average path length, high clustering coefficient, and power-law degree distribution [1], [2], [3]. Among of diverse network models, deterministic networks [4], [5], [6] have attracted increasing attention because they have an advantage with precise formulations on the above-mentioned attributes. In general, deterministic networks are defined against random networks where new nodes are connected to existing nodes with a certain probability. Fractal networks [7], [8], [9] constructed by fractal structures belong to a family of deterministic networks, the main issues that require consideration are network modeling [10], random walks [11], [12], and calculations of Laplacian spectrum [13].
Recently coherence in deterministic networks has been an emerging topic with aim on the interplay between network coherence and network topological structures [14], [15], [16]. Network coherence characterized by Laplacian spectrum of a network is used to measure the consensus dynamics with additive stochastic disturbances. For the consensus issue in the multi-agent systems, a key task is to design appropriate distributed consensus algorithms. There are many broad applications, such as satellite formation flying, cooperative unmanned vehicles, and sensor networks [17], [18], [19]. For the coherence of a network, the research goal is to study the effect of topological measurements on the coherence. Young et al. derived analytical expressions for first-order coherence in rings, path graphs and star graphs [14]. Patterson and Bamieh studied first and second order coherence in fractal networks and found that the scalings of network coherence are related to fractal dimension [15], [16]. In our recent published paper [20], it has been shown that the scalings of network coherence in a family of treelike networks do not depend on their fractal dimension, which shows that it is not a unique factor on the coherence behavior.
In the above-mentioned work [15], [16], the fractal dimensions of studied fractal networks are finite. A natural question arises, what are the scalings of coherence in a network with infinite fractal dimension. Therefore, we investigate the network coherence in the web graphs with infinite fractal dimension, which are defined as joining the pendant points of a helm to form a cycle and then adding a single pendant edge to each vertex of this outer cycle [21]. Here we propose a method to obtain exact expressions of first and second order coherence and obtain their scalings with network size N are N and . Obviously the scalings are not relevant to the fractal dimension. We also find that the scalings of web graphs are larger than those of other studied networks. Finally we investigate the effect of network parameter on the coherence behavior.
The rest of this paper is organized as follows. Section 2 gives the definitions of network coherence and constructions of web graphs. Calculations of network coherence are presented in Section 3. Section 4 includes the conclusions and discussions.
Section snippets
Definitions
Network coherence characterizes the variance of fluctuations in first and second order consensus systems with the additive noise [15]. It has been shown that the network coherence could be determined by the spectra of Laplacian matrix [14], [22], [23]. Let the eigenvalues of Laplacian matrix be , then the first and second order network coherence areand
Web graphs
The web graphs are formed by a positive-integer parameter after g generations.
Calculations of network coherence
In this section, we mainly calculate the network coherence characterized by Eqs. (1), (2). We need obtain the expressions for the sum of reciprocals of all nonzero eigenvalues and the square reciprocals of these eigenvalues.
Conclusions
In this paper, we have proposed a method to calculate first and second order network coherence quantifying as the sum of reciprocals and square reciprocals of all nonzero Laplacian eigenvalues. We then have obtained the scalings of network coherence in web graphs with infinite fractal dimension and compared them with those of other studied networks with finite fractal dimensions. We also have showed that the scalings of coherence in our networks do not depend on the fractal dimension and see
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Nos. 61203155, 11171084 and 11447005), and Zhejiang Provincial Natural Science Foundation of China under Grant No. LY13F030016.
References (25)
- et al.
Deterministic scale-free networks
Phys A
(2001) - et al.
Deterministic small-world communication networks
Inf Process Lett
(2000) - et al.
A deterministic small-world network created by edge iterations
Phys A
(2006) - et al.
Weighted fractal networks
Phys A
(2010) - et al.
Second-order consensus in multi-agent dynamical systems with sampled position data
Automatica
(2011) - et al.
Distributed average consensus with least-mean-square deviation
J Parallel Distr Com
(2007) - et al.
Statistical mechanics of complex networks
Rev Mod Phys
(2002) - et al.
Evolution of networks
Adv Phys
(2002) The structure and function of complex networks
Rev Mod Phys
(2003)- et al.
Apollonian networks: simultaneously scale-free, small world, Euclidean, space filling, and with matching graphs
Phys Rev Lett
(2005)
The rigorous solution for the average distance of a Sierpinski network
J Stat Mech
Novel evolving small-world scale-free Koch networks
J Stat Mech
Cited by (24)
Searching efficiency of multiple walkers on the weighted networks
2020, Physica A: Statistical Mechanics and its ApplicationsCitation Excerpt :These practical problems motivate us to find a general transition probability matrix on the basis of (HMF) theory and the annealed network approach. At the same time, for the properties of the network, previous researches mainly focused on the binary networks [24–29]. The weighted networks are not only the expansion of the binary networks, but also have a closer relationship with our real world [30–32].
Generalized adjacency and Laplacian spectra of the weighted corona graphs
2019, Physica A: Statistical Mechanics and its ApplicationsCitation Excerpt :Besides, the product graphs and their subgraphs (Hamming graphs, hypercubes, prisms and other classes of graphs) have a variety of applications which include problems from coding theory, frequency assignment, and mathematical chemistry [11]. However, previous researches mainly focused on the binary networks [12–17]. Weighted networks which are more in line with the real-world network are the extension of the binary networks.
First-order network coherence in 5-rose graphs
2019, Physica A: Statistical Mechanics and its ApplicationsCitation Excerpt :Zhan et al. studied asynchronous consensus problems of multiple double-integrator agents with discontinuous information transmission [6]. Sun et al. studied consensus problems in a family of tree networks and investigated first- and second-order consensus denoted as network coherence characterized by Laplacian spectrum [7,8]. Tyler et al. showed that network coherence is a submodular function of the network topology [5].
Mean first-passage times for two biased walks on the weighted rose networks
2019, Physica A: Statistical Mechanics and its ApplicationsCitation Excerpt :For example, in the area of communication and information networks, random walks can not only describe information delivery [1] and data gathering [2,3], but also predict the throughput [4,5], latency performance [1], transition [6] and search costs [7,8]. Other related applications of random walks include community detection [9,10], computer vision [11], network coherence [12], sampling networks [13,14], and so on. The statistical properties of random walks play an significance role in their applications, since they not only characterize the behaviour of random walks themselves, but also capture the performance metrics of different applications.
Network coherence and eigentime identity on a family of weighted fractal networks
2018, Chaos, Solitons and FractalsCitation Excerpt :Network coherence quantifies the steady-state variance of these fluctuations and can be considered as a measure of robustness of the consensus process to the additive noise. Sun et al. [8,9] studied the first order and second order network coherences and showed their scalings with network size N are lnN and N, respectively. Moreover, several recent works studied the mean first-passage time (MFPT) for some self-similar weighted networks.