AGC-based SU Selection for Cooperative Spectrum Sensing with Direct-Conversion Cognitive Radio

A direct-conversion receiver for cooperative spectrum sensing (CSS) was recently proposed to assess the performances of eigenvalue-based detectors and the energy detector (ED). In this paper, the problem of secondary user (SU) selection for CSS is solved by a classification method based on the automatic gain control loop gains of the receivers. Low complexity detectors are compared using this method: the Gerschgorin radii and centers ratio, the Gini index detector, the Pietra-Ricci index detector, and the locally most powerful invariant test. The ED is also included as a benchmark. Significant performance improvements are achieved in comparison with the use of the same number of SUs without selection.