复杂网络中基于近视最优反应的合作行为

Cooperative behavior in complex networks based on myopic best response

通过引入社会个体的行为一致性需求,在博弈个体的近视最优反应学习规则下探讨复杂网络中基于囚徒困境博弈的合作行为.分析表明,当个体邻居集合中的合作者多于背叛者,且合作成本与行为一致性需求的比率足够小时,规则网络上的合作行为能在个体的近视最优反应规则下得以演化,但存在导致群体演化为全面背叛均衡的临界比率条件.基于方格网和规则小世界网络的仿真实验验证了理论分析的结果,并发现小世界特性对合作行为的促进作用.用异质小世界网络的仿真实验表明,群体演化为全面背叛均衡的临界比率条件随小世界参数的增大而减小,合作水平随合作成本与行为一致性需求比率的减小而逐渐增大.

By introducing the behavioral conformity of social individuals,this paper explores the cooperative behaviors based on prisoner＇s dilemma in complex networks under the myopic best response of players.The theoretical analysis shows that the cooperative behaviors on regular networks can evolve under the myopic best response of players,when there are more cooperators than betrayer in neighbors set of players and the ratio of cooperative cost to the behavioral conformity is little enough.However,there are the critical ratios which lead to complete betrayal equilibrium.The simulation experiments based on square lattice and regular small world networks verified the results of theoretical analysis,and found the positive effect of small-world property on cooperative behaviors.Moreover,the simulation experiment based on heterogeneous small world networks shows that the critical ratios which lead to complete betrayal equilibrium decrease with the increase of small-world parameter,and the cooperative level increases with the decrease of the ratio of cooperative cost to the behavioral conformity.