Dynamic Threshold Correction based on the Exact Statistics of Energy Detection in Spectrum Sensing

Abstract

Cognitive radio is considered as a promising technology for efficient spectrum utilization. Spectrum sensing algorithm detects primary transmission based on the signal energy at the secondary receiver, well above a prefixed threshold under the binary hypothesis testing setup. In this paper, we investigate the performance of the well-known energy detector (ED), considering the distribution of the exact test statistics. We also consider the Gaussian approximation via the central limit theorem. We evaluate probability of overall error for ED, and derive an optimal threshold under the exact and Gaussian approximated statistics with noise variance uncertainty. Furthermore, we apply a dynamic correction mechanism to the optimal threshold with exact statistics of ED to compute the probability of error with noise variance uncertainty. Additionally, through Monte Carlo simulations, we show that the dynamic correction and threshold optimization significantly reduces the probability of error.