ABSTRACT
Abstract In MANETs, cooperative authentication, requiring the cooperation of neighbor nodes, is a significant authentication technique. However, when nodes participate in cooperation, their location may easily be tracked by misbehaving nodes; meanwhile, their resources will be consumed. These two factors lead to selfish nodes, reluctant to participate in cooperation, and will decrease the probability of correct authentication. To encourage nodes to take part in cooperation, we propose a bargaining-based dynamic game model for cooperative authentication to analyze dynamic behaviors of nodes and help nodes decide whether or not to participate in cooperation. Further, to analyze the dynamic decision-making of nodes, we discuss two situations: complete information and incomplete information, respectively. Under complete information, subgame perfect Nash equilibriums are obtained to guide nodes to choose their optimal strategy to maximize their utility. In reality, nodes often do not have good knowledge about others’ utility (this case is often called “incomplete information”). To deal with this case, the perfect Bayesian Nash equilibrium is established to eliminate the implausible equilibriums. Based on the model, we designed two algorithms for complete information and incomplete information, respectively, and the simulation results demonstrate that, in our model, nodes participating in cooperation will maximize their location privacy and minimize their resource consumption with an increased probability of correct authentication. Both the algorithms can improve the success rate of cooperative authentication and extend the network lifetime to 160%–360.6% of the present value.
