Local Information Promotes Cooperation in Duplex Public Goods Games with Limited Resources

Traditional works of public goods game focus on the situation that agents play the games in simplex networks in which social interactions are of the same type. However, due to multiple types of interactions in reality, agents always have to allocate the limited resources according to different interactions, such as the agents allocate the limited time to work with colleagues or accompany family members. Hence, in order to model the above situation, we study the public goods game in duplex networks which is named as duplex public goods game (D-PGG). In the duplex network, each agent has two types of social interactions and the network could be regarded as two network layers. In the D-PGGs, agents choose public goods in one network layer to allocate their limited resources. Besides, in the duplex network setting, local information is introduced by means of the single-layer information that agents could acquire from their neighbors in the corresponding network layer. Theoretical analyses and simulations conducted indicate that if agents imitate based on local information of the payoff, cooperation is promoted, and the extent of the promotion depends on the network structure and the similarity between different network layers. It is anticipated that our work can provide further insights to the effects of information on cooperation between agents in social networks.