SOCP based robust detector for cooperative spectrum sensing in MIMO cognitive radio

In this paper we present a robust detection scheme for cooperative spectrum sensing in cognitive radio (CR) networks with channel uncertainties. We consider a soft-decision scenario at the fusion center for primary user detection, based on the sensed test statistics submitted by the cooperating secondary users. The scheme presented models the channel state information (CSI) uncertainty employing an ellipsoidal uncertainty set. It is then demonstrated that the optimal linear discriminator for cooperative spectrum sensing towards primary user detection in a CR system can be formulated as a second order cone program (SOCP). Further, we also formulate a relaxed robust detector (RRD) and a multicriterion robust detector (MRD) that maximally separate the hypothesis ellipsoids at low signal-to-noise ratio (SNR) and deep fade conditions. Simulation results demonstrate that the detection error performance of the proposed CSI uncertainty aware robust spectrum sensing schemes are significantly lower compared to other uncertainty agnostic schemes.