Estimation of bias in numerical values of normalized sixth-order cumulants’ structures for various signal constellations

Abstract

Automatic modulation classification (AMC) represents technique for the process of modulation format recognition of signals considered to be a priori unknown, which finds wide usage in a numerous wireless systems and applications. Due to the low algorithm complexity and other practical aspects, AMC algorithms based on higher-order statistics are very popular. In this paper, properties of AMC algorithms based on sixth-order cumulants were explored in context of real and complex signals, i.e. different Pulse Amplitude Modulation (PAM) and Quadrature Amplitude Modulation (QAM) signal constellations. This was done through Monte Carlo simulations in a propagation conditions of Additive White Gaussian Noise (AWGN), but also in the channel with multipath propagation, with main focus on numerical quantification of the presence of bias – an interesting effect reported in standard cumulant structures. One new, recently published, approach in AMC, where unbiased nature of classification features is reported, was tested in the same context. Detailed numerical estimations for both standard and unbiased AMC algorithms are presented, confirming that bias issues can be resolved with new sixth-order cumulants formula and thus consequently better performances in classification of real and complex signals can be ensured, in all considered propagation scenarios.