Spectrum sensing plays a central role in cognitive radio (CR), because it represents the way that the radio looks for unused (and thus available) portions of the spectrum, often called spectrum holes. The detection of spectrum holes requires wideband spectrum sensing techniques, capable to observe a wide frequency band and to correctly identify which portions of such band contain only noise. In this work, we propose a novel wideband spectrum sensing technique relying on a model order selection problem, which in turn is solved by information theoretic criteria (ITC). This strategy exploits jointly spectral correlation and received energy to discern signals from noise in a blind way. Indeed the proposed algorithm does not require neither the knowledge of the noise power nor any a-priori information about the number, and the characteristics, of the signals to be detected. Numerical results reveal that the algorithm proposed outperforms the existing ones, especially in the presence of spectral correlation.
A. Giorgetti, A. Mariani, M. Chiani (2011). Spectrum holes detection by information theoretic criteria. NEW YORK, NY : ACM [10.1145/2093256.2093322].
Spectrum holes detection by information theoretic criteria
GIORGETTI, ANDREA;MARIANI, ANDREA;CHIANI, MARCO
2011
Abstract
Spectrum sensing plays a central role in cognitive radio (CR), because it represents the way that the radio looks for unused (and thus available) portions of the spectrum, often called spectrum holes. The detection of spectrum holes requires wideband spectrum sensing techniques, capable to observe a wide frequency band and to correctly identify which portions of such band contain only noise. In this work, we propose a novel wideband spectrum sensing technique relying on a model order selection problem, which in turn is solved by information theoretic criteria (ITC). This strategy exploits jointly spectral correlation and received energy to discern signals from noise in a blind way. Indeed the proposed algorithm does not require neither the knowledge of the noise power nor any a-priori information about the number, and the characteristics, of the signals to be detected. Numerical results reveal that the algorithm proposed outperforms the existing ones, especially in the presence of spectral correlation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.