Méthodes Cyclostationnaires d’Ordre Supérieur pour l’Estimation et la Détection de Signaux Numériques/ Higher-Order Cyclostationary Methods for the Estimation and Detection of Digital Signals.

Presenter: Vincent Gouldieff

Abstract: The advent of cutting-edge telecommunications techniques, such as intelligent radio (CR), has resulted in an increasing need for blind signal processing methods (spectrum sensing, standard detection, constellation recognition, intrinsic parameter estimation of digital signals). In this context, so-called “cyclostationary” methods have been studied since the 1970s and have rapidly shown a clear advantage over “classical” methods. In this presentation, we first propose some didactic reminders on the theory of cyclostationary estimation. In a second part, we apply these concepts in the case of blind recognition of constellations, especially in the advanced case of multi-user channels. Finally, we present some theoretical aspects related to the cyclostationary detection of digital signals in the context of spectrum sensing.