Date Added: Mar 2011
In the companion paper, a fast estimator for the capacity of a secondary communication in the context of cognitive radio networks was proposed. It was particularly shown that the proposed method largely outperforms traditional ones when the available number of samples is limited. In this paper, the authors study the fluctuations of the aforementioned estimators around their deterministic equivalents. They prove that in the asymptotic regime, their behaviors can be approximated by Gaussian random variables for which they derive the variances.