Detection Technology Spectrum At Low Signal To Noise Ratio

Spectrum detection is one of the important technologies in cognitive radio. Inactual communication environment, due to the influence of all kinds of fading,signal-to-noise ratio is very low. The Federal Communication Commission (FCC)regulates that spectrum detection works at least in-18dB. In order to ensure thesecurity of the cognitive radio network and maximize the protection to primary user,the spectrum detection in low snr can really be helpful to real application of cognitiveradio network.This paper firstly researched spectrum detection in low SNR, through seekingthe optimum threshold, designed an energy detection method，which can meetdetection time and detection performance requirement required by IEEE802.22standard. Then through using multi-user collaboration fusion method, verified themulti-user collaborative detection can reduce the sensing time, improve thedetection performance.Afterward, this paper researched spectrum detection technology based onspectral correlation, according to the low accuracy of periodogram algorithm, gavean improved spectral correlation detection algorithm based on MVDR(MinimumVariance Distortionless Response) spectrum. The simulation results show that thegiven method is better than the existing algorithm based on periodogram in sensingperformance for both signals, and through the analysis of the complexity of thealgorithm, we got this algorithm can satisfy the IEEE802.22requirement to detectiontime.Finally, according to the problem that the spectral correlation detectionalgorithm based on periodogram can’t solve the spectrum detection in multiplecognitive systems, both considering the detection probability and sensing time, wegave a spectral correlation detection algorithm based on Levison-Durbin algorithm.The simulation results show that, under the condition of multiple cognitive users, thedetection performance of this algorithm is slightly lower than the algorithm based onMVDR, but this algorithm can satisfy the requirement of real-time detection better.