An introduction to ROC analysis
Pattern Recognition Letters - Special issue: ROC analysis in pattern recognition
Analysis of area under the ROC curve of energy detection
IEEE Transactions on Wireless Communications
Hi-index | 0.00 |
In this article, we derive closed-form expressions for the area under the receiver operating characteristics (ROC) curve (hereafter, referred to as AUC metric) of energy detectors with/without diversity receiver over a myriad of fading environments. Our "unified" expression for the AUC performance metric only requires the knowledge of the derivatives of the moment generating function (MGF) of fading signal-to-noise ratio (SNR) at the receiver output. Aside from developing a new analytical framework for evaluating the AUC of energy detectors, we show that our approach also facilitates the treatment of a more generalized/realistic propagation environment (i.e., not restricted to Rayleigh/Nakagami-m channels). Moreover, the AUC analysis enables comparisons between various overlapping ROC curves to be performed easily. Different factors affecting the spectrum sensing performance were also analyzed and our key results and findings are useful in the design of cognitive radio networks.