Low Complexity Compressive Wideband Spectrum Sensing in Cognitive Radio

The capability of cognitive radio (CR) is realized by spectrum sensing. However, with the increase of signal bandwidth and the complexity of communication environment, traditional spectrum sensing has been facing a considerable challenge. Cyclic spectrum sensing techniques work well under noise uncertainty, but also require high-rate sampling. For realizing robust sub-Nyquist cyclostationary feature detection, we propose to reconstruct the conjugate cyclic spectrum of signals from a frequency domain representation at sub-Nyquist sampling rate. By investigating the link between the conjugate cyclic spectrum and the entries of the shifted conjugate correlation matrix, we transform the reconstruction of the cyclic spectrum into a solution to the correlation matrix. Further, a new reduced complexity method for reconstructing useful shifted conjugate correlation between frequency shifted versions of signals is presented. Simulations show that the proposed method has a high probability of detection and reconstruction accuracy against both noise uncertainty and sampling rate reduction.