Cyclostationary spectrum detection in cognitive radios

With the rapid development of wireless communication, it becomes more important to address the spectrum scarcity problem. In the licensed spectrum band, users only utilize their designated resources partially, thus necessitating the need of cognitive radios (CR) which offers the promising feature of accessing the unused spectrum by dynamic spectrum management. In this paper, we are presenting the cyclostationary spectrum density detection method for estimation and spectral autocorrelation function technique to analyze the spectrum. This technique is robust in a sense that it can detect active or licensed user signals blindly. For efficient spectrum detection Kaiser window function was used. The effect of the observational data length on signal detection is also included. A simulational analysis suggests that cyclostationary spectrum detection is optimal for signal detection having low signal-to-noise(SNR) values. For 10% false alarm probability, 90% detection probability of BPSK signals with SNR of -8 dB or greater was achieved.