Spectrum sensing for cognitive radio using multicoset sampling

Spectrum sensing is the very task upon which the entire operation of Cognitive Radio rests. In this paper, we propose a spectrum sensing technique based on the estimates of the spectrum of a multiband signal obtained from its non-uniform compressed multicoset samples. We show that our proposed spectrum sensing method provides accurate results using less data samples. We discuss in detail the effect of false detections on the quality of the reconstructed signal obtained from non-uniform multicoset samples.