Performance of Cyclostationary Features Based Spectrum Sensing Method in a Multiple Antenna Cognitive Radio System

In this paper, spectrum sensing methods in a multiple antenna cognitive radio (CR) system are discussed. We firstly present a simplified cyclostationary features based detector for single antenna, with relatively low computation complexity compared to traditional cyclostationary feature detection method. In order to achieve better signal detection performance in low SNR regions where the cyclostationary features are overwhelmed by strong noise, we resort to multiple antenna technique. We then propose three signal detection methods for multiple antenna scheme, namely maximum ratio combination method, multiple decision result fusion method and comparison detection method, which are motivated by MIMO techniques and structure characteristic of spectral correlation function estimation. The simulation results show the performance improvement of signal detection obtained by utilizing the three proposed multiple antenna processing methods, in terms of detection probability.