A Modified- Fast Spectrum Sensing with Coordinate System under Path Loss Effect and Noise Uncertainty for Cognitive Radio Networks

Abstract

Spectrum sensing is a critical function for cognitive radio which is used to continually monitor the activities of PU in the licensed spectrum band. In this paper, we first re-derive some parameters of FSC algorithm, which is called Modified- fast spectrum sensing with coordinate system (MFSC), to perform spectrum sensing under path loss effect and noise uncertainty. Then, performance of three knowledgebased spectrum sensing techniques — MFD, LED and MFSC — are evaluated under path loss effect and noise uncertainty. From the simulation results, we found that MFD gives the highest probability of detection when noise uncertainty does not exist. However its detection performance greatly degrades due to the occurrence of noise uncertainty. On the other hand, the effect of noise uncertainty does not cause any degradation to the detection performance of LED, however the detection performance of LED is worse than MFSC. When both factors — path loss and noise uncertainty — are both taken in the account, MFSC algorithm is the most achievable of spectrum sensing requirement since it gives high detection performance while consumes the least average sensing time.