A contribution to the characterization of the interference power in cognitive radio networks

In future cognitive radio networks, radio devices will observe by sensing to detect non-occupied bandwidth and interferences. The result of this observation process will then be used by the cognitive manager to adapt the transmitter to its environment, at all layers. The quality of the observation process is thus of primary importance. In this paper, we are interested in the statistical characterization of the power sensed by a radio node in the presence of multiple interferers when path-loss, shadowing and fast fading effects are taken into consideration. In this context, we develop a method that provides a very good approximation of the probability density function of this interference power, for a given localization of the interferers. In comparison with classical approaches that in general take into account path-loss and Rayleigh fading only, the proposed characterization also includes shadowing effects. Moreover, in addition to providing a better comprehension of the interference phenomenon when all propagation effects are included, the proposed model facilitates simulations compared to the classical Monte Carlo approach.