Noise analysis of limited length cyclostationary detection in cognitive radio systems

In order to improve the detection probability of weak signals,cyclostationary detection is used in Cognitive Radio system.However,the performance degradation for cyclostationary detection with limited data length can't be ignored,especially for noise signals.Cyclostationary spectrum estimation for noise signals with limited data length can greatly influence the performance of cyclostationary detection.The cyclostationary spectrum estimation of noise signals with limited data length(when cyclostationary frequency α≠0) was evaluated.Theoretical deduction and simulation results show that the noise spectral correlation function(when cyclostationary frequency α≠0) is proportion to 1/N(where N is the data length),which is essential for deducing the detection probability bound.