Optimal power allocation over parallel Gaussian channels in cognitive radio and jammer games

Cognitive jammers are able to deploy advanced strategies that degrade the performance of cognitive radio (CR) user communications. In this study, the authors study the problem of power allocation in CR user and jammer games, over parallel Gaussian channels. They model the interaction between a communicator (a transmitter-receiver pair) and a jammer using zero-sum games with continuous action sets; they describe unilateral, Nash and Stackelberg games. They compare the Nash equilibrium (NE), the Stackelberg equilibrium and the minmax/maxmin optimal power allocations through the simulation of the diverse game scenarios. Finally, they give the theoretical proof of existence and uniqueness of the NE.