1: \documentclass[11pt]{elsart}
2: \usepackage[dvips]{graphicx}
3:
4:
5: \newtheorem{property}{Property}
6: \setlength{\parindent}{8pt}
7: \begin{document}
8: \begin{frontmatter}
9:
10: \title{Growing Scale-free Small-world Networks with Tunable Assortative Coefficient
11: %\footnote{This work has been
12: %supported by the Chinese Natural Science Foundation of China under
13: %Grant Nos. 70431001 and 70271046.}
14: }
15:
16: \author{Qiang Guo\corauthref{jg}}
17: \corauth[jg]{Corresponding author. Tel. (+86)13050537943.}
18: \ead{liujg004@yahoo.com.cn}
19: \author{$^{a}$, Tao Zhou$^{b}$,}
20: %\ead{yzhdang@dlut.edu.cn}
21: \author{Jian-Guo Liu$^{c}$,}
22: \author{Wen-Jie Bai$^{d}$,}
23: \author{Bing-Hong Wang$^{b}$ and}
24: \author{Ming Zhao$^{b}$}
25: %\author{Jian-guo Liu}
26:
27: \address{$^{a}$ School of Science, DaLian Nationalities University,
28: Dalian 116600, P R China}
29: \address{$^{b}$ Department of
30: Modern Physics, University of Science and Technology of China, Hefei
31: Anhui, 230026, P R China}
32: \address{$^{c}$ Institute of System Engineering, Dalian University of Technology,
33: Dalian 116023, P R China}
34: \address{$^{d}$ Department of Chemistry, University of Science and Technology of China, Hefei
35: Anhui, 230026, P R China}
36:
37: %\title{$\Delta\Delta$ dibaryon structure in extended chiral
38: %SU(3) quark model}
39: %\thanks{Project supported by the Scientific
40: %Research Foundation of Liaoning Education Department (No.
41: %202122028) }
42: %\author{L. R. Dai}\thanks{Email: dailr@lnnu.edu.cn,
43: %Tel:0411-84259629}
44: %\affiliation{Department of Physics, Liaoning
45: %Normal University, 116029, Dalian, P. R. China}
46:
47:
48: \begin{abstract} \textnormal{\small {In this paper, we propose a
49: simple rule that generates scale-free small-world networks with
50: tunable assortative coefficient. These networks are constructed by
51: two-stage adding process for each new node. The model can reproduce
52: scale-free degree distributions and small-world effect. The
53: simulation results are consistent with the theoretical predictions
54: approximately. Interestingly, we obtain the nontrivial clustering
55: coefficient $C$ and tunable degree assortativity $r$ by adjusting
56: the parameter: the preferential exponent $\beta$. The model can
57: unify the characterization of both assortative and disassortative
58: networks.}}
59:
60: \begin{keyword}Complex networks, Scale-free networks, Small-world
61: networks, Assortative coefficient. \PACS 89.75.Da\sep 89.75.Fb\sep
62: 89.75.Hc
63: \end{keyword}
64: \end{abstract}
65: \date{}
66: \end{frontmatter}
67:
68: \section{Introduction}
69: \medskip
70: In the past few years, no issues in the area of network researching
71: attract more scientists than the ones related to the real networks,
72: such as the Internet, the World-Wide Web, the social networks, the
73: scientific collaboration and so on
74: \cite{WS98,BA99,AB02,DM02,New,XFWang01}. Recent works on the complex
75: networks have been driven by the empirical properties of real-world
76: networks and the studies on network dynamics
77: \cite{E1,E2,E5,EE5,E6,E7,E8,E9,E10,D1,D2,D3,D4,D5,D6,D7,D8,D9,D10,D11,D12,D13,D14}.
78: Many empirical evidences indicate that the networks in various
79: fields have some common topology characteristics. They have a small
80: average distance like random graphs, a large clustering coefficient
81: and power-law degree distribution \cite{WS98,BA99}, which are called
82: the small-world and scale-free characteristics. The other
83: characteristic is that the social networks are assortative while
84: almost all biological and technological networks are opposite. The
85: networks with high clustering and small average distance are the
86: small-world model of Watts and Strogatz (WS)\cite{WS98}, while the
87: networks with power-law degree distribution are the scale-free
88: network model of Barab$\acute{a}$si and Albert (BA) \cite{BA99}. The
89: BA model is a pioneering work in the studies on networks, which
90: suggests that the growth and preferential attachment are two main
91: self-organization mechanisms. Although BA model can generate the
92: power-law degree distributions, its assortative coefficient $r$
93: equals to zero in the limit of large size thus fail to reproduce the
94: disassortative property that extensively exists in the real-world
95: networks. Recently, some models that can generate either assortative
96: or disassortative networks have been reported
97: \cite{17,18,19,15,WWX2,WWX}. Wang {\it et al.} presented a mutual
98: attraction model for both assortative and disassortative weighted
99: networks. The model found that the initial attraction $A$ of the
100: newly added nodes may contribute to the difference of the
101: assortative and disassortative networks \cite{WWX2}. Liu {\it et
102: al.} \cite{cpl} proposed a self-learning mutual selection model for
103: weighted networks, which demonstrated that the self-learning
104: probability $p$ may be the reason why the social networks are
105: assortative and the technological networks are disassortative.
106: However, one should not expect the existence of a omnipotent model
107: that can completely illuminate the underlying mechanisms for the
108: emergence of disassortative property in various network systems. In
109: this paper, beside the previous studies, we exhibit an alternative
110: model that can generate scale-free small-world networks with tunable
111: assortative coefficient, which may shed some light in finding the
112: possible explanations to the different evolution mechanisms between
113: assortative and disassortative networks.
114:
115:
116: Dorogovtsev {\it et. al} \cite{M6} proposed a simple model of
117: scale-free growing networks . In this model, a new node is added to
118: the network at each time step, which connects to both ends of a
119: randomly chosen link undirected. The model can be equally described
120: by the process that the newly added node connect to node $i$
121: preferentially, then select a neighbor node of the node $i$
122: randomly. Holme and Kim \cite{M7} proposed a model to generate
123: growing scale-free networks with tunable clustering. The model
124: introduced an additional step to get the trial information and
125: demonstrated that the average number of trial information controls
126: the clustering coefficient of the network. It should be noticed that
127: the newly added node connect to the first node $i$ preferentially,
128: while connect to the neighbor node of the first node $i$ randomly.
129: In this paper, we will propose a growing scale-free network model
130: with tunable assortative coefficient. Inspired by the above two
131: models, the new node is added into the network by two steps. In the
132: first step, the newly added node connects to the existing nodes $i$
133: preferentially. In the second step, this node selects a neighbor
134: node $s$ of the node $i$ with probability $k_s^{\beta}/\sum_{j\in
135: {\Gamma_i}} k_j^{\beta}$, where $\beta$ is the parameter named
136: preferential exponent and $\Gamma_i$ is the neighbor node set of
137: node $i$. This model will be equal to the Holme-Kim model[37] when
138: $\beta=0$, and the MRGN model[35] when $\beta=1$ Specifically, the
139: model can generate a nontrivial clustering property and tunable
140: assortativity coefficient. Therefore, one may find explanations to
141: various real-world networks by our microscopic mechanisms.
142:
143:
144:
145:
146:
147: \section{Construction of the Model}
148: \medskip
149:
150: Our model is defined as following.
151:
152: \begin{description}
153: \item[(1)] Initial condition: The model starts with $m_{0}$
154: connected nodes.\\[2pt]
155:
156: \item[(2)] Growth: At each time step, one new node $v$ with $m$
157: edges is added at every time step. Time $t$ is identified as the
158: number of time steps.
159:
160: \item[(3)] The first step: Each edge of $v$ is then
161: attached to an existing node with the probability proportional to
162: its degree, i.e., the probability for a node $i$ to be attached to
163: $v$ is
164: \begin{equation}
165: \Pi_i=\frac{k_i}{\sum_j k_j}.
166: \end{equation}
167:
168: \item[(4)] The second step: If an
169: edge between $v$ and $w$ was added in the first step, then add one
170: more edge from $v$ to a randomly chosen neighbor $s$ of $w$ with
171: probability $P_t$ according to the following probability
172: \begin{equation}
173: p_{s}=\frac{k_{s}^{\beta}}{\sum_{v\in \Gamma_i}{k_{v}^{\beta}}},
174: \end{equation}
175: If there remains no pair to connect, i.e., if all neighbors of $w$
176: were always connected to $v$, do the first step instead.
177:
178: \end{description}
179:
180: \section{Characteristics of the Model}
181: \medskip
182:
183: \subsection{Degree distribution}
184: The degree distribution is one of the most important statistical
185: characteristics of networks. Since some real-world networks are
186: scale-free, whether the network is of the power-law degree
187: distribution is a criterion to judge the validity of the model. By
188: adopting the mean-field theory, the degree evolution of individual
189: node can be described as
190: \begin{equation}\label{F2.1}
191: \frac{\partial k_i}{\partial t}=P(i)+\sum_{j\in
192: \Gamma_i}P(i|j)P(j),
193: \end{equation}
194: where $P(i)$ denotes the probability that the node $i$ with degree
195: $k_i$ is selected at the first step, $P(i|j)$ denotes the
196: conditional probability that node $i$ is a neighbor of node $j$ with
197: degree $k_j$ which has been selected at the first step.
198:
199:
200: According to the preferential attachment mechanism of the first
201: step, one has
202: \begin{equation}\label{F2.2}
203: P(i)=\frac{k_i}{\sum_j k_j}.
204: \end{equation}
205: The conditional probability $P(i|j)$ can be calculated by
206: \begin{equation}\label{F2.3a}
207: P(i|j)=\frac{k_i}{\sum_{l\in \Gamma_j}k_l}.
208: \end{equation}
209: According to the second step, one has that
210: \begin{equation}\label{F2.5}
211: \frac{\partial k_i}{\partial t}=\frac{k_i}{\sum_l k_l}+\sum_{j\in
212: \Gamma_i}\Big(\frac{k_i^{\beta}}{\sum_{s\in \Gamma_j}
213: k_s^{\beta}}\frac{k_j}{\sum_l k_l}\Big).
214: \end{equation}
215: If $\beta=0$, we get that
216: \begin{equation}\label{F2.6}
217: \frac{\partial k_i}{\partial t}=\frac{k_i}{\sum_l k_l}+\sum_{j\in
218: \Gamma_i}\Big(\frac{1}{k_j}\frac{k_j}{\sum_l
219: k_l}\Big)=\frac{2k_i}{\sum_l k_l}.
220: \end{equation}
221: Then we can get that $P(k)\sim k^{-3}$, which has been proved by
222: Holme and Kim \cite{M7}. If $\beta=1$, the following formula can be
223: obtained under the assumption that the present network is
224: non-assortative.
225: \begin{equation}\label{F2.7} \frac{\partial
226: k_i}{\partial t}=\frac{k_i}{\sum_j k_j}+\sum_{j\in
227: \Gamma_i}\frac{k_i}{\langle k\rangle k_l}\frac{k_l}{\sum_j
228: k_j}=\frac{2k_i}{\sum_j k_j}.
229: \end{equation}
230: We can obtain that the degree distribution $p(k)\sim k^{-r}$ obeys
231: the power-law and the exponent $\gamma=3$. The numerical results are
232: demonstrated in Fig. 1. From Fig. 1, we can get that the exponents
233: $\gamma$ of the degree distribution are around -3 when
234: $\beta=-2,-1,0,1,2$. When $\beta>0$, the exponent $\gamma$ would
235: increase slightly as the $\beta$ increases.
236:
237:
238:
239:
240:
241:
242: \begin{figure}[ht]
243: \begin{center}
244: %{\includegraphics[0,0][270,180]{N=4.bmp}}
245: \center \includegraphics[width=13cm]{figure1.eps}
246: \caption{(Color Online) Degree distribution of the present network with $N=100000$ nodes when $m=3$ and $P_t=0.3$.
247: In this figure, $p(k)$ denotes the probability of nodes with
248: degree $k$ in the network. The power-law degree distribution exponents $\gamma$ of the
249: four probability density function are $\gamma_{\beta=2}=3.11\pm 0.05$,
250: $\gamma_{\beta=1}=3.11\pm 0.05$, $\gamma_{\beta=0}=2.93\pm 0.04$, $\gamma_{\beta=-1}=2.96\pm 0.05$ and
251: $\gamma_{\beta=-2}=2.95\pm 0.04$.}
252: \end{center}
253: \end{figure}%\label{F01}
254:
255:
256:
257: \begin{figure}
258: \begin{center}
259: %{\includegraphics[0,0][270,180]{N=4.bmp}}
260: \center \includegraphics[width=13cm]{figure2.eps}
261: \caption{(Color Online) The dependence between the average distance $D$ and
262: the order $N$ of the present network, when $\beta=1$ and $m=4$. One
263: can see that $L$ increases very slowly as ${\rm ln}N$ increases. The
264: inset exhibits the curve where $L$ is considered as a function of
265: $N$. All the data are obtained by 10 independent simulations.}
266: \end{center}
267: \end{figure}%\label{F01}
268:
269: \subsection{Average distance}
270:
271: The average distance is also one of the most important parameters
272: to measure the efficiency of communication networks, which is
273: defined as the mean distance over all pairs of nodes. The average
274: distance plays a significant role in measuring the transmission
275: delay. Firstly, we give the following lemma \cite{M1}.
276:
277: {\bf Lemma 1} {\it For any two nodes $i$ and $j$, each shortest
278: path from $i$ to $j$ does not pass through any nodes $k$
279: satisfying that $k>{\rm max}\{i,j\}$.
280: %\end{lemma}
281: }
282:
283: {\it Proof.} Denote the shortest path from the node $i$ to $j$ of
284: length $n+1$ by $i\rightarrow x^1 \rightarrow x^2 \cdots \rightarrow
285: x^n \rightarrow j $($SP_{ij}$), where $n>0$. Suppose that $x^k={\rm
286: max}\{x^1, x^2, \cdots, x^n\}$, if $x^k\leq {\rm max}\{i, j\}$, then
287: the conclusion is true. If $x^k>{\rm max}\{i,j\}$, denote the
288: youngest node of $SP_{ij}$ by $k$. Denote the subpath passing
289: through node $k$ by $l\rightarrow k\rightarrow r$, where the node
290: $l$ and $r$ are the neighbors of node $k$, then we can prove that
291: node $l$ and $r$ are connected. The shortest path $SP_{ij}$ passes
292: from the node $l$ to $r$ directly, which is conflicted with the
293: hypothesis.
294:
295: Let $d(i,j)$ represent the distance between node $i$ and $j$ and
296: $\sigma(N)$ as the total distance, i.e., $\sigma(N)=\sum_{1\leq
297: i<j\leq N}d(i,j)$. The average distance of the present model with
298: order $N$, denoted by $L(N)$, is defined as following
299: \begin{equation}\label{F3.1}
300: L(N)=\frac{2\sigma(N)}{N(N-1)}.
301: \end{equation}
302: According to Lemma 1, the newly added node will not affect the
303: distance between the existing ones. Hence we have
304: \begin{equation}\label{F3.2}
305: \sigma(N+1)=\sigma(N)+\sum^N_{i=1}d(i,N+1).
306: \end{equation}
307: Assume that the $(N+1)$th node is added to the edge $E_{y_1y_2}$,
308: then Equ. (\ref{F3.2}) can be rewritten as
309: \begin{equation}\label{F3.3}
310: \sigma(N+1)=\sigma(N)+N+\sum^N_{i=1}D(i,y),
311: \end{equation}
312: where $D(i,y)={\rm min}\{d(i,y_1),d(i,y_2)\}$. Denote $y$ as the
313: edge connected the node $y_1$ and $y_2$ continuously, then we have
314: the following equation
315: \begin{equation}\label{F3.4}
316: \sigma(N+1)=\sigma(N)+N+\sum_{i=\Lambda}d(i,y),
317: \end{equation}
318: where the node set $\Lambda=\{1,2, \cdots, N\}-\{y_1, y_2\}$ has
319: $(N-2)$ members. The sum $\sum_{i=\Lambda}d(i,y)$ can be considered
320: as the distance from each node of the network to node $y$ in the
321: present model with order $N-2$. Approximately, the sum
322: $\sum_{i=\Lambda}d(i,y)$ is equal to $L(N-2)$. Hence we have
323: \begin{equation}\label{F3.5}
324: \sum_{i=\Lambda}d(i,y)\approx (N-2)L(N-2).
325: \end{equation}
326: Because the average distance $L(N)$ increases monotonously with
327: $N$, this yields
328: \begin{equation}\label{F3.6}
329: (N-2)L(N-2)=(N-2)\frac{2\sigma(N-2)}{(N-2)(N-3)}<\frac{2\sigma(N)}{N-1}.
330: \end{equation}
331: Then we can obtain the inequality
332: \begin{equation}\label{F3.7}
333: \sigma(N+1)<\sigma(N)+N+\frac{2\sigma(N)}{N-1}.
334: \end{equation}
335: Enlarge $\sigma(N)$, then the upper bound of the increasing
336: tendency of $\sigma(N)$ reads
337: \begin{equation}\label{F3.8}
338: \frac{d\sigma(N)}{dN}=N+\frac{2\sigma(N)}{N-1}.
339: \end{equation}
340: This leads to the following solution
341: \begin{equation}\label{F3.9}
342: \sigma(N)=(N-1)^2{\rm log}(N-1)+C_1(N-1)^2-(N-1).
343: \end{equation}
344: The numerical results are demonstrated in Fig. 2.
345:
346:
347: \subsection{Clustering property}
348:
349:
350:
351: \begin{figure}
352: \begin{center}
353: %{\includegraphics[0,0][270,180]{N=4.bmp}}
354: \center \includegraphics[width=12cm]{figure3.eps}
355: \caption{(Color Online) The scale of $C$ with various $\beta$ and $P_t$ when $m=3$. The data are averaged over 10 independent runs of network size of $N=6000$.}
356: \end{center}
357: \end{figure}%\label{F01}
358: The small-world characteristic consists of two properties: large
359: clustering coefficient and small average distance. The clustering
360: coefficient, denoted by $C$, is defined as
361: $C=\sum_{i=1}^N\frac{C_i}{N}$, where $C_i$ is the local clustering
362: coefficient for node $i$. $C_i$ is
363: \begin{equation}\label{F4.1}
364: C_i=\frac{2E(i)}{k_i(k_i-1)},
365: \end{equation}
366: where $E(i)$ is the number of edges in the neighbor set of the
367: node $i$, and $k_i$ is the degree of node $i$. When the node $i$
368: is added to the network, it is of degree $m+mP_t$ and $E(i)=mP_t$.
369: If a new node is added to be a neighbor of $i$ at some time step,
370: $E(i)$ will increase by $mP_t$ since the newly added node will
371: connect with one of the neighbors of the node $i$ with probability
372: $P_t$. Therefore, in terms of $k_i$, the expression of $E(i)$ can
373: be written as
374: \begin{equation}\label{F4.2}
375: E(i)=mP_t+P_t[k_i-(m+mP_t)].
376: \end{equation}
377: Hence, we have that
378: \begin{equation}\label{F4.3}
379: C_i=\frac{2[mP_t+P_t(k_i-m-mP_t)]}{k_i(k_i-1)}=2\Big(\frac{mP_t^2}{k_i}+\frac{P_t-mP_t^2}{k_i-1}\Big).
380: \end{equation}
381: This expression indicates that the local clustering scales as
382: $C(k)\sim k^{-1}$, where $C(k)$ denotes the average clustering
383: coefficient value of nodes with degree $k$. It is interesting that a
384: similar scaling has been observed in many artificial models
385: \cite{M7,M1,M2,M8} and several real-world networks \cite{M9}. The
386: degree-dependent average clustering coefficient $C(k)$ has been
387: demonstrated in Fig. 4. Consequently, we have
388: \begin{equation}\label{F4.4}
389: C=\frac{2}{N}\sum^N_{i=1}\Big(\frac{mP_t^2}{k_i}+\frac{P_t-mP_t^2}{k_i-1}\Big).
390: \end{equation}
391: \begin{figure}
392: \begin{center}
393: %{\includegraphics[0,0][270,180]{N=4.bmp}}
394: \center \includegraphics[width=12cm]{figure4.eps}
395: \caption{(Color Online) $C(k)$ vs $k$ to various $\beta$ and when $m=3$ and $P_t=0.3$.
396: The data are averaged over 10 independent runs of network size of $N=6000$.}
397: \end{center}
398: \end{figure}%\label{F01}
399: Since the degree distribution is $P(k)=a k^{-3}$, where $k=k_{\rm
400: min},\cdots, k_{\rm max}$. The constant $a$ satisfies the
401: normalization equation
402: \begin{equation}
403: \sum_{k_{\rm min}}^{k_{\rm max}}ak^{-3}=1,
404: \end{equation}
405: one can get that $a=2k_{\rm min}^2$. The average clustering
406: coefficient $C$ can be rewritten as
407: \begin{equation}
408: \begin{array}{rcl}
409: C & = & \frac{2}{N}\sum_{k_{\rm min}}^{k_{\rm max
410: }}\Big(\frac{NP(k)mP_t^2}{k}+\frac{NP(k)(P_t-mP_t^2)}{k-1}\Big)\\[2pt]
411: & \approx & 2\sum_{k_{\rm min}}^{k_{\rm max
412: }}\Big(ak^{-4}mP_t^2+ak^{-3}(P_t-mP_t^2)/(k-1)\Big)\\[2pt]
413: %& = & 2\sum_{k_{\rm min}}^{k_{\rm max }}ak^{-4}P_t.
414: \end{array}
415: \end{equation}
416: The numerical results are demonstrated in Fig.3. From figure 3, we
417: can get that if $\beta\leq 0$, the numerical results are consistent
418: with the theoretical predictions approximately, while if $\beta>0$,
419: the fluctuations emerges. The departure from analysis results is
420: observed, which may attribute to the fluctuations of the power-law
421: exponent of degree distribution. It is also helpful to compare the
422: present method with previous analysis approaches on clustering
423: coefficient for Holme-Kim model \cite{AD1,AD2}.
424:
425: \subsection{Assortative coefficient}
426: The assortative coefficient $r$ can be calculated from
427: \begin{equation}
428: r=\frac{M^{-1}\sum_i j_i
429: k_i-[M^{-1}\sum_i\frac{1}{2}(j_i+k_i)]^2}{M^{-1}\sum_i\frac{1}{2}(j_i^2+k_i^2)-[M^{-1}\sum_i\frac{1}{2}(j_i+k_i)]^2},
430: \end{equation}
431: where $j_i$, $k_i$ are the degrees of the vertices of the $i$th
432: edge, for $i=1, 2, \cdots,M$ \cite{r,r2}.
433:
434: \begin{figure}
435: \begin{center}
436: %{\includegraphics[0,0][270,180]{N=4.bmp}}
437: \center \includegraphics[width=12cm]{figure5.eps}
438: \caption{(Color Online) The scale of $r$ with various $\beta$ and $P_t$ when $m=1$.
439: The data are averaged over 10 independent runs of network size of $N=6000$.}
440: \end{center}
441: \end{figure}%\label{F01}
442:
443: From Fig. 5, we can find that when $\beta>0$ the assortative
444: coefficient $r$ increases with the probability $P_t$, while $r$
445: decreases with the probability $P_t$ when $\beta<0$. As $\beta=0$,
446: $r$ equals to zero approximately.
447:
448:
449:
450: \section{Conclusion and Discussion}
451: In this paper, we propose a simple rule that generates scale-free
452: small-world networks with tunable assortative coefficient. The
453: inspiration of this model is to introduce the parameter $\beta$ to
454: Holme-Kim model. The simulation results are consistent with the
455: theoretical predictions approximately. Interestingly, we obtain the
456: nontrivial clustering coefficient $C$ and tunable degree
457: assortativity $r$, depending on the parameters $\beta$. The model
458: can unify the characterization of both assortative and
459: disassortative networks. Specially, studying the degree-dependent
460: average clustering coefficient $C(k)$ also provides us with a better
461: description of the hierarchies and organizational architecture of
462: weighted networks. Our model may be conducive to future
463: understanding or characterizing real-world networks.
464:
465:
466:
467:
468:
469:
470:
471:
472:
473:
474:
475:
476: This work has been supported by the Chinese Natural Science
477: Foundation of China under Grant Nos. 70431001, 70271046 and
478: 70471033.
479:
480:
481:
482: \begin{thebibliography}{WS98}
483: \bibitem{WS98}
484: D. J. Watts and S. H. Strogatz,
485: % Collective dynamics of¡®small-world¡¯ networks,
486: Nature 393 (1998) 440.%-442.
487:
488: \bibitem{BA99}
489: A. -L. Barab$\acute{a}$si and R. Albert,
490: %Emergence of scaling in random networks, Science,
491: Science 286 (1999) 509.%-512.
492:
493: \bibitem{AB02}
494: R. Albert and A. -L. Barab\'{a}si,
495: %Statistical mechanics of complex networks,
496: Rev. Mod. Phys. 74 (2002) 47.%-97.
497:
498: \bibitem{DM02}
499: S. N. Dorogovtsev and J. F. F. Mendes,
500: %Evolution of networks,
501: Adv. Phys. 51 (2002) 1079.%-1187.
502:
503: \bibitem{New}
504: M. E. J. Newman,
505: %The structure and function of complex networks,
506: SIAM Rev. 45 (2003) 167.%-256.
507:
508: %\bibitem{DM00} S.N. Dorogvtsev, J.F.F. Mendes,
509: %Scaling behaviour of developing and decaying networks,
510: %Europhys. Lett. 52 (2000) 33.%-39.
511:
512:
513: \bibitem{XFWang01}
514: X. F. Wang, Int. J. Bifurcat. Chaos 12 (2002) 885.
515:
516: %\bibitem{XFWangC02}
517: %X. -F. Wang and G. -R. Chen 2003 {\it IEEE Circuits Systems Mag.}
518: %{\bf 3} 6
519:
520: \bibitem{E1}
521: W. Li and X. Cai, Phys. Rev. E 69(2004) 046106.
522:
523: \bibitem{E2}
524: R. Wang and X. Cai, Chin. Phys. Lett. 22 (2005) 2715.
525:
526: %\bibitem{E3}
527: %Chi L P et al %, B. Wang, H. Su, X. -P. Xu, J. -S. Zhao, W. Li and X. Cai
528: %2003 {\it Chin. Phys. Lett.} {\bf 20} 1393
529:
530: %---------------------he da-ren
531: %\bibitem{E4}
532: %Xu T et al
533: %%, J. -C. Chen, Y. He and D. -R. He
534: %2004 {\it Int. J.
535: %Mod. Phys. B} {\bf 18} 2599
536:
537: \bibitem{E5}
538: P. -P. Zhang, K. Chen, Y. He, T. Zhou, B. -B. Su, Y. -D. Jin, H.
539: Chang, Y. -P. Zhou, L. -C. Sun, B. -H. Wang and D. -R. He, Physica A
540: 360 (2005) 599.
541:
542: \bibitem{EE5}
543: P. -P. Zhang, Y. He, T. Zhou, B. -B. Su, H. Chang, Y. -P. Zhou, B.
544: -H. Wang and D. -R. He, Acta Physica Sinica 55 (2006) 60.
545: %--------------------Di Zeng-ru
546: \bibitem{E6}
547: M. H. Li, Y. Fan, J. -W. Chen, L. Gao, Z. -R. Di and J. -S. Wu,
548: Physica A 350 (2005) 643.
549:
550: \bibitem{E7}
551: J. Q. Fang and Y. Liang, Chin. Phys. Lett. 22 (2005) 2719.
552:
553:
554: \bibitem{E8}
555: F. C. Zhao, Y. J. Yang and B. -H. Wang, Phys. Rev. E 72 (2005)
556: 046119.
557:
558: \bibitem{E9}
559: H. J. Yang, F. C. Zhao, L. Y Qi and B. L. Hu, Phys. Rev. E 69 (2004)
560: 066104
561:
562: \bibitem{E10}
563: J. -G. Liu, Y. -Z. Dang and Z. -T. Wang, arXiv: physics/0509183.
564:
565: \bibitem{D1} A. E. Motter and Y. -C. Lai, Phys. Rev. E 66 (2002)
566: 065102.
567:
568: \bibitem{D2} K. -I. Goh, D. -S. Lee, B. Kahng and D. Kim, Phys. Rev. Lett.
569: 91 (2003) 148701.
570: \bibitem{D3} T. Zhou and B. -H. Wang, Chin. Phys. Lett. 22 (2005)
571: 1072.
572:
573: \bibitem{D4} T. Zhou, B. -H. Wang, P. -L. Zhou, C. -X. Yang and J. Liu,
574: Phys. Rev. E 72 (2005) 046139.
575:
576: \bibitem{D5} M. Zhao, T. Zhou, B. -H. Wang and W. -X. Wang, Phys. Rev. E 72 (2005)
577: 057102.
578:
579: %\bibitem{D13} Zhou T et al 2005 {\it Preprint} arXiv:
580: %cond-mat/0508368.
581:
582: \bibitem{D6} W. Q. Duan, Z. Chen, Z. R. Liu and W. Jin
583: Phys. Rev. E 72 (2005) 026133.
584:
585: \bibitem{D7} F. Jin, L. Xiang and F. W. Xiao, Physica A 355 (2005)
586: 657.
587:
588: %\bibitem{O1}
589: %A. X. C. N. Valente, A. Sarkar and H. A. Stone, Phys. Rev. Lett. 92
590: %(2004) 118702.
591:
592: %\bibitem{O2}
593: %G. Paul, T. Tanizawa, S. Havlin and H. E. Stanley, Eur. Phys. J. B
594: %38 (2004) 187.
595:
596: \bibitem{D8}
597: B. Wang, H. W. Tang, C. H. Guo, Z. L. Xiu and T. Zhou,
598: {\it Preprint} arXiv:cond-mat/0509711.
599:
600: \bibitem{D9}
601: B. Wang, H. W. Tang, C. H. Guo and Z. L. Xiu, {\it Preprint}
602: arXiv:cond-mat/0506725.
603:
604: \bibitem{D10}
605: J. -G. Liu, Z. -T. Wang and Y. -Z. Dang, Mod. Phys. Lett. B 19
606: (2005) 785.
607:
608: \bibitem{D11}
609: J. -G. Liu, Z. -T. Wang and Y. -Z. Dang, {\it Preprint}
610: arXiv:cond-mat/0509290.
611:
612: \bibitem{D12}
613: C. P. Zhu, S. J. Xiong, Y. J. Tian, N. Li and K. S. Jiang, Phys.
614: Rev. Lett. 92 (2004) 218702
615:
616: \bibitem{D13}
617: F. Comellas, G. Fertin and A. Raspaud, %Recursive graphs with
618: %small-world scalefree properties,
619: Phys. Rev. E 69 (2004) 037104.
620:
621: \bibitem{D14}
622: F. Comellas and M. Sampels, %Deterministic small-world networks,
623: Physica A 309 (2002) 231.%-235.
624:
625:
626:
627:
628:
629:
630: \bibitem{17}
631: A. V\'{a}zquez, Phys. Rev. E {\bf 67}, 056104 (2003).
632:
633: \bibitem{18}
634: R. Xulvi-Brunet and I. M. Sokolov, Phys. Rev. E 70 (2004) 066102.
635:
636: \bibitem{19}
637: M. Catanzaro, G. Caldarelli and L. Pietronero, Phys. Rev. E 70
638: (2004) 037101.
639:
640: \bibitem{15}
641: A. Barrat, M. Barth\'{e}lemy, R. Pastor-Satorras and A. Vespignani,
642: Proc. Natl. Acad. Sci. U.S.A. 101 (2004) 3747.
643:
644:
645: \bibitem{WWX2}
646: W. -X. Wang, B. Hu, T. Zhou, B. -H. Wang and Y. -B. Xie, Phys. Rev.
647: E 72 (2005) 046140.
648:
649: \bibitem{WWX}
650: W. -X. Wang, B. -H. Wang, B. Hu, G. Yan and Q. Ou, Phys. Rev. Lett.
651: 94 (2005) 188702 .
652:
653:
654: \bibitem{cpl}
655: J. -G. Liu, Y. -Z. Dang and Z. -T. Wang, Chin. Phys. Lett. 23(3)
656: (2006) 746.
657:
658: \bibitem{M6}
659: S. N. Dorogovtsev, J. F. F. Mendes1 and A. N. Samukhin, Phys. Rev. E
660: 63 (2001) 062101.
661:
662: \bibitem{M7}
663: P. Holme and B. J. Kim, Phys. Rev. E 65 (2002) 026107.
664:
665: \bibitem{M1}
666: T. Zhou, G. Yan and B. -H. Wang, Phys. Rev. E 71 (2005) 046141.
667:
668: \bibitem{M2}
669: J. S. Andrade, H. J. Herrmann, R. F. S. Andrade and L. R. da Silva,
670: Phys. Rev. Lett. 94 (2005) 018702.
671:
672: \bibitem{M8}
673: S. N. Dorogovtsev, A. D. Goltsev and J. F. F. Mendes, %Pseudofractal
674: %scale-free web,
675: Phys. Rev. E 65 (2002) 066122.
676:
677: \bibitem{M9}
678: E. Ravasz and A. -L. Barab$\acute{a}$si,
679: %Hierarchical organization in complex networks,
680: Phys. Rev. E 67 (2003) 026112.
681:
682:
683: \bibitem{AD1}
684: G. Szab\'{o}, M. Alava and J. Kert\'{e}sz, Phys. Rev. E 67 (2003)
685: 056102.
686:
687: \bibitem{AD2}
688: X. Wu, B. -H. Wang, T. Zhou, W. -X. Wang, M. Zhao and H. -J. Yang,
689: Preprint arXiv: cond-mat/0511655 (Chin. Phys. Lett. In Press).
690:
691:
692:
693: \bibitem{r}
694: M. E. J. Newman, Phys. Rev. Lett. 89 (2002) 208701.
695:
696: \bibitem{r2}
697: M. E. J. Newman, Phys. Rev. E 67 (2003) 026126 .
698:
699:
700:
701:
702:
703:
704:
705:
706:
707:
708:
709: %\bibitem{D1} B. Tadi\'{c}, S. Thurner and G. J. Rodgers, Phys. Rev. E 69 2004
710: %036102.
711:
712:
713: %\bibitem{D2} L. Zhao, Y. C. Lai, K. Park and N. Ye, Phys. Rev. E 71 (2005)
714: %026125.
715:
716: %\bibitem{D3} Yan G et al 2005 {\it Preprint} arXiv:
717: %cond-mat/0505366.
718:
719: %\bibitem{D4} Yin C Y et al 2005 {\it Preprint} arXiv:
720: %physics/0506204.???
721:
722: %\bibitem{D5} R. Pastor-Satorras and A. Vespignani, Phys. Rev, Lett. 86 (2001)
723: %3200.
724:
725: %\bibitem{D6} Yan G et al 2005 {\it Chin. Phys. Lett.} {\bf 22} 510.
726:
727: %\bibitem{D7} Zhou T et al 2005 {\it Preprint} arXiv: physics/0508096
728:
729:
730:
731:
732:
733: %\bibitem{M5}
734: %Zhang Z Z and Rong L L 2005 {\it Preprint} arXiv:cond-mat/0502591
735:
736:
737:
738:
739:
740:
741:
742: %\bibitem{M10}
743: %Jiang P Q et al 2005 {\it Chin. Phys. Lett.} {\bf 22} 1285
744:
745:
746:
747:
748:
749: %\bibitem{New89}
750: %M. E. J. Newman, Phys. Rev. Lett. 89 (2002) 208701.
751: \end{thebibliography}
752: \end{document}
753: