Efficient Dynamic Double Threshold Energy Detection of Cooperative Spectrum Sensing in Cognitive Radio

Abstract

With the increasing number of wireless users every day, cognitive radio serves as an approach to solve the spectrum crunch problem. Spectrum sensing serves as the heart of cognitive radio with it being able to decide if an unlicensed user is to be given access to a licensed band or not without causing interference. Various spectrum sensing techniques have been discussed in the literature. In this paper, energy detection is considered for spectrum sensing, which conventionally works on a fixed threshold without requiring any prior knowledge of the signal. It is found the performance of conventional energy detection falls in regions with noise uncertainty. In this paper, a dynamic double threshold scheme along with cooperative spectrum sensing at the fusion center is proposed. The dynamic threshold selection works on the parameter of noise uncertainty for practical cases by creating a noise variance history. Also, the fusion center uses a dynamic threshold to make a final decision compared to a fixed threshold for all the energy values lying between the two thresholds at the local nodes. A simulation model has been discussed to compare the proposed scheme with traditional energy detection and other detection schemes as well. A 20% improvement in probability of detection at −22 dB SNR and 0.5 probability of false alarm is achieved using the proposed scheme.