Distributed coalition formation for selfish relays and eavesdroppers in wireless networks: A job-hopping game

Game theory has been widely applied to modeling cooperative or competitive behaviors among wireless users and their adversaries in physical layer security. In this paper, we analyze a novel cooperative behavior among eavesdroppers and relays to enhance their eavesdropping performance. The above cooperative eavesdropping problem is formulated as a hedonic coalition formation game, where we propose a novel and well suited job-hopping preference for the game players (i.e. the eavesdroppers and jamming relays). Simulation results show that the game players can self-organize into an ultimate coalition structure through a distributed hedonic job-hopping algorithm. Besides, it is clear that our scheme can improve the average eavesdropping capacity per eavesdropper up to 88.33% over the non-cooperative case. Also, by autonomous and dynamic job-hopping behaviors, the players can self-adapt the network topology to inherent environmental changes in wireless networks.