Corrected John’s Test Based Blind Spectrum Sensing Technique for Cognitive Radio Networks

Abstract

Spectrum sensing is one of the fundamental objective of cognitive radio network (CRN). In last one-decade the eigenvalue-based blind spectrum sensing methods have been extensively studied for CR applications. Eigenvalue-based techniques require large data sample for accurate detection of the signal under the low Signal-to-Noise ratio. This results in delayed sensing under large data sample. Thus, the recent trend is to explore the eigenvalue-detection methods under the low sample environment.In this paper we propose a Corrected John’s Test (CJT) based eigenvalue technique for spectrum sensing. Asymptomatic test statistic and Probability of false alarm are obtained for the same. Performance analysis of CJT detector is compared with the previously proposed techniques like, Scaled Largest Eigenvalue (SLE), Maximum-Minimum Eigenvalue (MME), and Arithmetic to Geometric Mean (AGM) for relatively less number of samples. The simulation results show supremacy of CJT under Nakagami fading environment, for very less number of samples.