Procedure for automatic recognition of analogue and digital modulations

Abstract
A global procedure for recognition of analogue and digital modulation types, utilising the decision-theoretic approach, is developed. It involves two main steps: a reliable classification of the signal according to its amplitude and phase information contents; and a classification of the signal according to the modulation type. Computer simulations for different types of band-limited modulated signals corrupted with band-limited Gaussian noise sequence have been carried out. It is found that the threshold SNR for correct signal classification is about 10 dB.