Spectrum sensing using sub-nyquist rate sampling

Spectrum sensing in wideband regime requires huge amount of samples. The observed frequency spectrum is usually sparse. Compressed sensing technique provides a viable solution to reconstruct the sparse signals. The observed wideband spectrum can be reconstructed using compressive sensing technique. Inherent constraints of the compressed sensing algorithms hinder the flexible implementation of spectrum sensing process. The structure-based Bayesian sparse recovery algorithm is used in this paper to implement spectrum sensing process. Spectrum sensing performed using the Bayesian estimation approach resulted in better performance compared to the results based on compressed sensing technique. Various cases have been discussed considering the amount of information available for the observed frequency band. Spectrum sensing performed using the Bayesian algorithm showed improvement of more than 5 dB in all cases.