Spectrum sensing using higher-order statistics and receive diversity and cooperative detection in SIMO cognitive radio system

In this paper, we present a new spectrum sensing method with high-order statistics (HOS) and receive diversity for improved sensing performance in single- input multiple-output (SIMO) cognitive radio (CR) system. The detection method based on HOS is a blind detection method and has the advantage of suppressing Gaussian noise. In a SIMO CR system, a primary user (PU) is considered as the transmitter with one antenna and a secondary user (SU) is considered as the receiver equipped with multiple antennas. The secondary user detects if a licensed band is occupied by the primary user. The spectrum sensing method proposed combines the third-order statistics (bispectrum) with receive diversity in the SIMO CR system. Furthermore, we consider cooperative spectrum sensing that is regarded as the centralized detection problem with false alarm constraint in multi-user CR system. OR decision fusion rule is applied. Simulation results demonstrate that the proposed spectrum sensing method performs well in the range of low SNR and benefits from the receive diversity of the SIMO CR system. Multi-user cooperative spectrum sensing can further improve the performance of the signal detection of the system.