A Comprehensive Framework for Spectrum Sensing in Non-Linear and Generalized Fading Conditions

We derive a comprehensive analytical framework for the ED over generalized, extreme, and non-linear fading conditions which addresses the topic completely. This is carried out for both conventional and diversity receptions and it is based on the area under the ROC curve (AUC), which is an efficient performance measure that is widely used in physical sciences and engineering. This differentiates the considered methodology from the aforementioned routine approaches and additionally provides generic results on the arbitrary derivatives of the MGF of useful generalized processes. The asymptotic behavior of the derived expressions is also analyzed providing direct and concrete insights on the role and effect of the involved parameters on the ED performance. The offered analytic results are subsequently employed in quantifying the performance of ED over various types of fading conditions, which exhibits that ED performance is significantly degraded by even slight variations of the severity of fading. To this end, it is shown that the detrimental effects of fading can be effectively mitigated with the aid of square-law combining and switch-and-stay combining methods, as a low number of diversity branches can ensure sufficient and holistic performance improvement even in severe fading conditions.