LAUSR

Abstract Cooperation can be enhanced via adaptive local interactions with a range of parameters, i.e., if the weakest neighbor of a player has more than one neighbor, then this player severs its connections with their weakest neighbor and randomly establishes a link to one of the neighbors of the weakest neighbor. We find that the degree of most nodes is small after the temporary station, while only a few nodes having large degrees. Most cooperative agents only have a small number of neighbors and link to players with small degrees. However, the defecting players occupy locations with large degrees and prefer to link to players with large degrees. The cooperators constitute a giant component of the network, and there is no local cluster for all values of b ; the defectors constitute a giant component from the local clusters with defection temptation b . A larger noise parameter κ suppresses the individuals from reconnecting γ and strategy changing ω and promotes the spreading of cooperation. Players in a uniform lattice structure are very likely to change their neighbors and are unlikely to change their strategies, whereas players in a scale-free (SF) network are very likely to change their strategies and unlikely to change their neighbors. Specifically, the players in an Erdős–Renyi (ER) network are an intermediate case.