international workshop on complex networks, seoul (23-24 june 2005) vertex correlations,...

International Workshop on Complex Networks, Seoul (23-24 June 2005)

Vertex Correlations, Self-Avoiding Walks and Critical Phenomena on the Static

Model of

Scale-Free Networks

DOOCHUL KIM (Seoul National University)

Collaborators:

Byungnam Kahng (SNU), Kwang-Il Goh (SNU/Notre Dame), Deok-Sun Lee (Saarlandes), Jae- Sung Lee (SNU), G. J. Rodgers (Brunel) D.H. Kim (SNU)


Outline

I. Static model of scale-free networks

II. Vertex correlation functions

III. Number of self-avoiding walks and circuits

IV. Critical phenomena of spin models defined on the static model

V. Conclusion


I. Static model of scale-free networks

static model


static model

We consider sparse, undirected, non-degenerate graphs only.

1. Degree of a vertex :

2. Degree distribution:

,1

N

i jj

aik=

=å

( ) l-:DP k k

G= adjacency matrix element

,i ja


static model

For theoretical treatment, one needs to take averages over an ensemble of graphs G. Grandcanonical ensemble of graphs:

( ) ( )=åG

O P G O G

Static model: Goh et al PRL (2001), Lee et al NPB (2004), Pramana (2005, Statpys 22 proceedings), DH Kim et al PRE(2005 to appear)

Precursor of the “hidden variable” model [Caldarelli et al PRL (2002), Soederberg PRE (2002) , Boguna and Pastor-Satorras PRE (2003)]


static model

1. Each site is given a weight (“fitness”)

2. In each unit time, select one vertex with prob. and another vertex with prob. .

3. If or already, do nothing (fermionic constraint).Otherwise add a link, i.e., set .

4. Repeat steps 2,3 times (= time = fugacity = ).

1

( 1, , ), 1, (0 1)m m m- -

=

= = = < <å åLN

i iij

P i j i N P

Construction of the static model

= Zipf exponent

When =0 ER case. Walker algorithm (+Robin Hood method) constructs

networks in time . network in 1 min on a PC.C

om

men

ts


static model

Such algorithm realizes a “grandcanonical ensemble” of graphs ={} with weights

1( ) (1 ) -

Î Ï >

= = -Õ Õ Õ ij ija aij ij ij ij

b G b G i j

P G ff ff(1- )

2=0 1 2 1-= - = = -i jK NP PNKij i j ija P P fProb( ) ( ) e

2=1 1 -= = - i jK NP Pij ija fProb( ) e

Each link is attached independently but with inhomegeous probability f i,j .


static model

1

2 2 2 2

1 1

( ) Poissonian with mean 2 ,

/ 2 ,

/ ,

1( ) Scale-free with 1

i i i

N

ii

N N

i ii i

D

P k k KNP

k k N K

k k N k k NP

P k

Degree distribution

Percolation Transition

2

2

( 1)( 3)for 31

2( 2)2

0 for 3C

ii

KNP


static model

Recall

When (),

When ()

Com

men

ts

ln

ln

iN

ln

ln

jN

Strictly uncorrelated in links, but vertex correlation enters (for finite N) when 2<<3 due to the “fermionic constraint” (no self-loops and no multiple edges) .

21 -= - i jK NP Pijf e

, 2 /i j i j i jf KNPP k k k N

1/( 1)max 1 ~k k N


II. Vertex correlation functions

Vertex correlations

nn ( ) average degree of nearest neighbors of vertices with

degree (ANND).

( ) clustering coefficient of vertices with degree

k k

k

C k k

Related work: Catanzaro and Pastor-Satoras, EPJ (2005)


2.5

2.5

3 4

3 4

Vertex correlations

Simulation results of k nn (k) and C(k) on Static Model


Vertex correlations

Our method of analytical evaluations:

, ,, , , ,,, ,

nn

( 1)( 1) ( 1)

( )i j j mi j j m i j j m

j i m i jj i m i j j i m i j

i i i

a aa a f f

k ik k k

11 1 ( ) ( ) ( ) ( ) (1)

N NN

mF x dx F N F m F x dx F

For a monotone decreasing function F(x) ,

1/ 2( , ~ , ~ / )j iN N x k NP k N

( 1),

,

( 1) (1 )

(1 )

xyj m

m i jx

f e y dy

O e

Use this to approximate the first sum as

Similarly for the second sum.


3 2

1 1

23 (3 ) 1

for ~( ) ~

for ~

cnn

c

N k k Nk k

N k k k N

Vertex correlations

Result (1) for( ) 2< <3nnk k


2

2

( 2 8 7) /( 1) (3 ) 1/ 2

5 2 2(3 ) 1/ 2 1/( 1)max

ln for

( ) ~ for

for ~

C

C

N N k k

C k N k k k N

N k N k k N

Vertex correlations

Result (2) for

( ) 2< <3C k


iC010

-210

-410

-610-810

310

710

510

Result (3) Finite size effect of the clustering coefficient for 2<<3:

Vertex correlations

2

ln ( ) ~

NC C i

N

1c.f. for 3C

N


III. Number of self-avoiding walks and circuits

Number of SAWs and circuits

The number of self-avoiding walks and circuits (self-avoiding loops) are of basic interest in graph theory.

Some related works are: Bianconi and Capocci, PRL (2003), Herrero, cond-mat (2004), Bianconi and Marsili, cond-mat (2005) etc.

Issue: How does the vertex correlation work on the statistics for 2<<3 ?


Number of SAWs and Circuits

The number of L-step self-avoiding walks on a graph is

' '

, , , , , ,L i j j k l m i j j k l ma a a f f f N

where the sum is over distinct ordered set of (L+1) vertices,{ , , , }.i j m We consider finite L only.

The number of circuits or self-avoiding loops of size L on a graph is

' '

, , , , , ,L i j j k l i i j j k l ia a a f f f M

with the first and the last nodes coinciding.



1

, , , ,11( , , ) 1

( 1) (1 ) (1 )xy x

N N

i j i j i t ij i m j

e y dy e

f f f dt f

1 2 3 4

0.2

0.4

0.6

0.8

1

Strategy for 2< <3: For upper bounds, we use

repeatedly. Similarly for lower bounds with

The leading powers of N in both bounds are the same.

Note: The “surface terms” are of the same order as the “bulk terms”.

1

0(1(1 ) ) ( ) x

ee L x

1/ 2( , ~ , )i

N N x k N P

0

if 0 11 ( )

1 if 1x x x

e L xx

and



Result(3): Number of L-step self-avoiding walks

3 For , straightforward in the static model

For , the leading order terms in N are obtained.

2 3

2

( 1)2

( 1)

(3 )31 1

2( 1)2

31 1 ( 1) / 22

1 ( 3)

~ 0.25 ln ( 3)

~ (2 3, even)

~ ln (2 3, odd)

L

L

L

L

L

L

L L

L

kN k

k

N k k N

N N L

N N L

N

N

N

N



Typical configurations of SAWs (2<λ<3)

L=2

L=3

L=4

1/( 1) 0H H

22 H H

~ , ~

~

k N n N

k n

N

1/ 2 (3 ) / 2S S

2 2 2 23 S S H H

~ , ~

~

k N n N

k n k n

N

( 2) /( 1) (3 )SS SS

2 24 H SS SS

~ , ~

~

k N n N

k k n

N

H

H H

S S

B B

BB

B BSS



( )

2

(3 ) / 2

(3 ) / 2

11 ( 3)

2

~ 0.25 ln ( 3)

~ ln (2 3, even)

~ (2 3, odd)

L

L

L

L

LL

LL

k

L k

k N

N N L

N L

M

M

M

M

-

-

æ ö< > ÷ç ÷= - >ç ÷ç ÷ç< >è ø

< > =

< < =

< < =

Results(4): Number of circuits of size

( 3)L L


IV.Critical phenomena of spin models defined on the static model

Spin models on SM


Spin models on SM

Spin models defined on the static model network can be analyzed by the replica method.

, , ,[ ] with quenched bond disorder i j i j i j i ji j

H G a J S S a

QQQQQQQQQQQQQQQQQQQQQQQQQQQQ

0

, , , eff

Free energy in replica method= log Z lim 1 /

tr (1 ) exp( ) tr exp[ ]

n

n

ni j i j i j i j

i j

Z n

Z f f J S S H


eff , ,

,

ln 1 exp( ) 1

exp( ) 1 ( )

i j i ji j

i j i j i ji j

H f J S S

k NPP J S S o N




Spin models on SM

3

2

( ln ) for 2 3

( ) ~ ((ln ) ) for 3

(1) for 3

O N N

o N O N

O

The effective Hamiltonian reduces to a mean-field type one with infinite number of order parameters,

1 1 1

, , ,N N N

i i i i i i i i ii i i

P S P S S P S S S

QQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQ

When J i,j are also quenched random variables, do additional averages on each J i,j .


Spin models on SM

We applied this formalism to the Ising spin-glass [DH Kim et al PRE (2005)]

, , ,( ) ( ) (1 ) ( )i j i j i jP J r J J r J J

Phase diagrams in T-r plane for 3 and<3


Spin models on SM

Critical behavior of the spin-glass order parameter in the replica symmetric solution:

1

1/( 3)

1 2 2

2( 2) /(3 )

( ) ( 4)

( ) / ln( ) ( 4)

~ ( ) (3 4)

exp( 2 / ) ( 3)

(2 3)

C

C CN

i i i Ci

T T

T T T T

q P S S T T

T T k

T

2/(3 )21

1~ ~ for 2 3

N

i ii

qQ S S T

N k T

( 2) /(3 ) 1/(3 )

1 1

1~ , ~

N N

i i ii i

m P S T M S TN

To be compared with the ferromagnetic behavior for 2<<3;


V. Conclusion1. The static model of scale-free network allows detailed

analytical calculation of various graph properties and free-energy of statistical models defined on such network. 2. The constraint that there is no self-loops and multiple links introduces local vertex correlations when , the degree exponent, is less than 3.

3. Two node and three node correlation functions, and the number of self-avoiding walks and circuits are obtained for 2< <3. The walk statistics depend on the even-odd parity.

4. The replica method is used to obtain the critical behavior of the spin-glass order parameters in the replica symmetry solution.

international workshop on complex networks, seoul (23-24 june 2005) vertex correlations,...

Documents