Comparison of cyclostationary and energy detection in cognitive radio network

In cognitive radio network (CRN) three types of spectrum sensing techniques are widely used: matched filter detection, cyclostationary detection, and energy detection. Matched filter detection technique provides maximum SNR at receiving end but the detector need to be matched with the input signal i.e. impulse response of the system need to be made as the delayed version of mirror image of input signal. If the modulation scheme and window function of baseband signal of primary user (PU) is not known to a secondary user (SU) then it is difficult to use matched filter detection. In this paper we only consider cyclostationary detection and energy detection of their simplicity compared to matched filter detection. Main focus of the paper is to measure the probability of successful access and corresponding delay of a SU in CRN. We found that energy detection technique is better compared to cyclostationary detection.