Analysis of Improved Cyclostationary Detector with Multiple Antennas over Fading Channels

Abstract

A comprehensive performance analysis of the multi-cycle cyclostationary (MC) detection-based spectrum sensing over fading channels with multiple independent and correlated antennas is developed. We first proposed an improved MC detector, aiming to reduce the computational complexity of the conventional one. Compared with conventional MC detector, the proposed method is low-computational complexity and high-accuracy on sensing performance. Based on the proposed MC detector, for the multiple independent antennas case, the average detection probability by employing square-law combining (SLC) is derived for several fading channels such as Nakagami, Rayleigh and Rician by using the moment generation function (MGF) approach. For multiple correlated antenna case, with Nakagami fading and SLC scheme, expressions of detection probability are derived by the same approach as it in the independent antennas case. Special cases of a linear array of 2 and 4 arbitrarily correlated antennas are treated. Finally, illustrative and analytical results show that the reliability of our proposed MC detector and the degradation of sensing performance over correlation and fading.