Estimators of the Nakagami-m parameter and performance analysis

Abstract

The Nakagami-m parameter is known to capture the envelope distribution of various fading channel conditions in wireless communications. The value of m is indicative of the severity of fading, and is a measure of channel quality, making its estimation necessary in many applications. In this paper, we summarize the existing estimators for the Nakagami-m parameter and propose a new class of estimators, whose performance is analyzed by deriving the asymptotic variance and comparing with the Cramer-Rao bound (CRB). Moreover, we develop a novel estimator robust to the presence of additive white Gaussian noise (AWGN) or any other additive noise with a symmetrical distribution. We also discuss practical issues not previously addressed in the literature including adaptation and computational complexity of these estimators. We conclude that our novel integer-moment based estimators are the best choice from a computational complexity and performance point of view. Simulation results corroborate our analysis.