Strategies to Promote Cooperation in Mobile Networks

In self-organized systems such as mobile ad hoc networks, decisions taken by individual nodes may damage the interests of the whole system. Mobile users need to cooperate with each other to improve the whole system’s payoffs. We address the problem by constructing a random mobile multi-agent system, in which agents playing repeated Prisoner’s Dilemma game and Snowdrift game with each other. Agents can remember the game history and make decisions. There are two evolutionary modes: combat–offspring–diffusion (COD) and combat–diffusion– offspring (CDO). The simulation results show that CDO mode can promote cooperation; memory can promote cooperation especially for the Prisoner’s dilemma game. We also studied the R, ST, P reciprocity and strategies distribution of the system. For the Prisoner’s dilemma game, "Win-stay-lose-shift" is the dominant strategy; For the Snowdrift game, "1011" is the dominant strategy, which promotes the ST reciprocity.