Evolutionary Games And Evolutionary Cooperation Dynamics In Complex Networks

After a multi-year voyage around the world, Charles Robert Darwin, the British biologist and the founder of the evolution theory, published his world-shaking and landmark masterwork, On the Origin of Species, on the basis of field survey and specimen collection. In his writings, he puts forward the famous evolution rule “natural selection and survival of the fittest”. Many of his ideas provide the subsequent studies much inspiration and guidance. According to Darwinian selection, individuals with higher fitness are always much favored by selection in the process of evolution and then selfish behavior will finally becomes widespread in the population. In sharp contrast, altruists who often help others via incurring a cost to themselves are expected to die out, which undoubtedly conflicts with the extensive altruism. Up to now, this evolution problem has aroused considerable attention of scholars from various fields. Evolutionary game theory, together with complex network theory, provides a powerful mathematical framework for addressing the issues concerning the evolution of altruism. This thesis mainly focuses on problems regarding the evolution of altruism in structured populations, containing the following contents:First, we study the effects of recommendation on the evolution of cooperation in spatial public goods games. Each group can recommend their members for establishment of new relationships with individuals outside the current group according to group quality(i.e., fraction of cooperators of the group). Main results show that for ER graphs and square lattices of periodic boundary conditions there exists the optimal recommendation threshold that can induce positive feedback between cooperation and recommendation, whereby the level of both cooperation and recommendation can be simultaneously maximized. Local recommendation turns out to be more favorable to the evolution of cooperation compared to global recommendation. Besides, results for BA scale free networks show that significant heterogeneity of BA scale free networks weakens the differences that different recommendation thresholds bring about in terms of promoting cooperation.Second, we also investigate the role of population mobility in the evolution of cooperation with the prisoner?s dilemma game and snowdrift game, respectively. Then we first examine effects of migration on diluted lattices and check the influences of various preferences of individuals toward living environments on the evolution of cooperation. It is shown that cooperation prevails well when individuals are inclined to migrate to a richer environment, i.e., when individuals frequently aspire for interactions with more successful ones. Soon afterwards, based on the fact that previous studies seldom distinguish mobility on the individual level, we further explore the impacts of payoff-related velocity on cooperation. We find that relatively low mobility can strikingly boosts cooperation provided the challenge to cooperate is not too serious. In contrast, high mobility always inhibits cooperation. In addition, the evolution of cooperation seemingly rises superior to selection strength once the mobility intensity is fixed in this case.Moreover, we analyze the coupled dynamics of strategies and topologies both in two-person games and in multi-person games. Adaptive dynamical linking based on the feedback from the game is considered. We find that cooperation can be remarkably enhanced by this feedback based co-evolutionary scheme both for the game of pairwise interactions, such as the prisoner?s dilemma game, and for the game of group interactions, illustrated by the public goods game. And the faster the adjusting rate, the more successful the evolution of cooperation. Additionally, aspiration-based partner switching is also studied. In this scenario, individuals adjust social ties in a self-questioning manner in line with learning theory. More specifically, each individual determines whether or not to modulate topologies based on the relationship between their aspirations and actual payoff rather than based on the strategies or performance of their rivals. We find that intermediate aspirations most favor the evolution of cooperation. And it turns out to be quite beneficial to the evolution of altruistic behavior when individuals prefer to maintain relationships with those of high payoff during the topology reorganizing process.Finally, in consideration of wide existence of various preferences and asymmetric role allocations of social agents in the real social activities, we check the evolution of social fairness in the spatial ultimatum game in which payoff-based partner selection and reputation-based role assignation rule are embedded. Our results indicate that fairness can be dramatically promoted if agents are willing to interact with more successful players and meanwhile tend to designate reputable ones as proposers in the ultimatum bargaining.