Behavior Dynamics in Cognitive Radio Networks: An Interacting Particle System Approach

A key feature of cognitive radio network is the intelligence of secondary users which can collaborate to improve the performance. The collaboration in terms of channel recommendation is studied. Via recommendations, the channel preferences of secondary users become dynamic. The corresponding behavior dynamics of secondary users are studied. Particularly, the ergodicity of the dynamical progress is studied using the model of interacting particles in nonequilibrium statistical mechanics, which is important for predicting the long term dynamics of the cognitive radio network. A mean field description using ordinary differential equation for the dynamics is also used to explicitly describe the behavior dynamics. A Jackson network with lattice topology is simulated to demonstrate the conclusions.