Centralized cooperative spectrum sensing with multiple antennas over imperfect feedback channels

In this paper, we study the performance of a cooperative spectrum sensing (CSS) scheme over imperfect feedback channels, where N secondary users (SUs) collaborate in order to enhance the detection probability of a primary user (PU). The sensing process is divided into two phases: In the first phase each SU, which is equipped with multiple antennas, independently detect the PU signal. In the second phase, the SUs' local decisions are sent via erroneous and orthogonal feedback channels to a fusion center (FC), where a final decision is made. We utilize Bayesian criterion which accounts for the costs of misdetection and false alarm. Quasi-Newton method is employed to obtain the optimal number of antennas for a given SNR and vice versa. Our simulation results concur the mathematical derivations and validate the optimality of quasi-Newton method. We further show that the derived optimal fusion rule outperforms the OR rule, AND rule and MAJORITY rule.