Performance of spectrum sensing algorithms under fading, electromagnetic interference and outliers

In this paper, we study the effects of shadowing-fading, electromagnetic interference and outliers on sequential algorithms in detecting spectral holes in a Cognitive Radio set up. The statistics of the primary signal, channel gain and the EMI are not known. Different nonparametric sequential algorithms are compared to choose appropriate algorithms to be used for energy detection and mean change detection. Modification of a recently developed random walk test is found to work well for energy detection as well as for mean change detection. We show via simulations and analysis that the nonparametric algorithms developed are robust to fading, EMI and outliers.