Widely Linear Cooperative Spectrum Sensing for Cognitive Radio Networks
Angela Sara Cacciapuoti, Marcello Caleffi, Luigi Paura · 2010
The Spectrum Sensing is a fundamental functionality of the Cognitive Radio (CR) paradigm since a reliable detection of the presence of spectral holes is a mandatory requirement to avoid harmful interferences to licensed users. In this paper, we address this problem by proposing a widely linear (WL) cooperative strategy at a fusion center side where the CR user statistics have been received by means of multipath fading channels. The proposed strategy jointly elaborates the received signal and its conjugate version to take advantage by the information contained in the statistical pseudo-covariance function of the received signal. We design the WL strategy to maximize a modified deflection coefficient and we compare it with a strictly linear (SL) cooperative strategy. Results of numerical simulations show that the WL fusion-rule outperforms the SL one in operative conditions of practical interest.