Exploiting Sparse Property of the Cyclic Autocorrelation Function for Cyclostationary Process Compressed Sensing

Ziad Khalaf (1) , Jacques Palicot (1)

1 IETR - Institut d'Électronique et des Technologies du numéRique

Based on the use of compressed sensing applied to recover the sparse cyclic autocorrelation (CA) in the cyclic frequencies domain, this paper proposes a new blind method based on compressed sensing in order to recover the CA. This new estimation method uses only few samples (short observation time) in order to estimate the CA. This proposed blind method outperforms the classical unbiased, non blind estimator used in [12], in term of estimation error, evaluated by calculating the Mean Square Error (MSE) compared to the theoretical equation of the CA. This result is validated by simulation from the theoretical case that does not include neither noise nor a filter at the transmission, to the most realistic case that includes a transmission filter and by adding noise at the reception side.

Keywords: Cognitive radio; Spectrum sensing; Sparsity; Compress Sensing.; Compress Sensing

Domains: Computer Science [cs]; Signal and Image Processing; Engineering Sciences [physics]; Signal and Image processing

Ziad Khalaf, Jacques Palicot. Exploiting Sparse Property of the Cyclic Autocorrelation Function for Cyclostationary Process Compressed Sensing. WSR 2012, Mar 2012, Karlsruhe, Germany. 6 p. ⟨hal-00682669⟩