Complex Network-Based Study On Organization Model

The complex network has obtained the plentiful research results after many yearsof development. It gets researchers’more and more attention who come from varioussubject areas. Many systems which are in the social and economic life such as Internet,WWW and rail transportation network can be described by the complex network. Inthe complex network, nodes representing objects in the system and the edges whichcontact between nodes represent the object. Evolutionary game theory comes fromDarwin’s theory of evolution and it is often used in researching group’s behavior.Assume that individuals in the group are of bounded rationality. Bounded rationalindividuals in the game wanted to profit the most in the current period and they areinclined to betray strategy. However, in actual widespread cooperation exists in thesystem. For a long time, Game theory is considered to be the most powerful means ofresearch on cooperation, and complex networks provide a powerful research tool forevolutionary game theory. Nodes in the network represent individuals and edgesbetween nodes represent relation between individuals. In a network, individuals canonly to play with its neighbors. This mechanism is closer to the truth, to depict actualgame system. Many studies have shown that, high levels of cooperation in largepopulations can be achieved in an environment where individual and interactionstructures co-evolve. Cooperative behavior in a group can greatly enhance thecompetitiveness of the entire group as well as of the individual. The research ofindividual behaviors and interaction structures co-evolve and its application in thesocial and economic life will gets more attention.This paper introduce an network growth mechanism on the basis of individualbehaviors and interaction structures co-evolve mechanism and study its impact on theemerging of cooperation and the structure of network. Because of that Snow-Driftgame is more conducive to the survival of the cooperation strategy, so Snow-Drift isselected in the paper as the game types between players in the network. The model inthe paper uses the rule proposed by Traulsen etc which has a universal form with the probability of strategy replication as the rule which the individual renew its strategywith. The level of cooperation, degree heterogeneity, clustering coefficient and degreecorrelations of the four indicators are studied by computer simulation. This paper getsthat the level of cooperation in growth network reduces through the analysis ofsimulation results. The network structure is also changed. Degree heterogeneity anddegree correlations strengthened and clustering coefficient weakened. Because ofmany factors are concerned, analysis in mathematics is very difficult and traditionalanalysis methods could not meet the requirements. We apply multivariate regressiontechniques in analysis the impact of various factors on the four indicators. The level ofcooperation in the model built in this paper is lower than the level in the model whosescale is fixed. In this state, the proportion of cooperators in the network is more stable.The main innovation of this paper: existing research considering individuals’self-organization on evolutionary game model are mostly concentrated in theinteraction between the network’s spatial topology dynamic change and individualstrategy replication, and the influence on cooperative behavior emerging. This type ofevolution model is only considered network whose scale is fixed. This paper proposesa network growth mechanism. In this mechanism, this paper studies the influence ofindividual’s self-organization on the network topology and the emergence ofcooperation.