An information-jamming-based incentive mechanism for channel selection

Abstract

We consider a radio cell composed of one single Base Station (BS) and possibly several users. We are particularly interested in analyzing the behavior of two special users. On one hand, the first user (called the sender) has access to multiple channels and has to choose a channel to transmit its data. Obviously, the sender seeks to transmit using the best channel among the L available channels in the system. Indeed, he/she “senses” each channel by sending probes to the BS. Then, the base station sends back the global interference (interference plus noise) on each channel. On the other hand, the second user (called the jammer) blocks the information sent by the BS to the user for some channels. We formulate this interaction as a zero-sum non-cooperative game. Then, we discuss the optimal strategy for each user and analyze the equilibrium point of the game for pure and mixed strategies. Many results are drawn such as characterization of pure strategies and derivation of saddle point policies for both players. Next, we provide a discussion on how our proposal could be implemented for cooperative sensing in cognitive radio networks. Our solution has the benefit of being completely distributed and near optimal channel assignment is met without coordination/signaling. We finally show, by solving the derived dynamic programs, that setting a transmission cost sustains fairness and incites for better channel assignment.