Publication detail

Gaussian Mixture Models-based Recognition of Digital Modulations of Noisy Signals

KUBÁNKOVÁ, A. ATASSI, H. KUBÁNEK, D.

Original Title

Gaussian Mixture Models-based Recognition of Digital Modulations of Noisy Signals

English Title

Gaussian Mixture Models-based Recognition of Digital Modulations of Noisy Signals

Type

journal article - other

Language

en

Original Abstract

A novel approach to the recognition of digital modulations is presented. It is based on features that characterize a specific kind of modulation or a group of modulations. These features are calculated from the instantaneous amplitude, instantaneous phase, and spectrum symmetry of the unknown signal. To recognize the modulations, an analysis of the features is carried out by means of a classifier based on Gaussian mixture models. The method was designed for 2ASK, 2FSK, MSK, BPSK, QPSK, and 16QAM modulations. They belong to the widely used digital modulations in recent communication systems. Simulation results demonstrating performance of the method are presented. The testing signals are corrupted by white Gaussian noise.

English abstract

A novel approach to the recognition of digital modulations is presented. It is based on features that characterize a specific kind of modulation or a group of modulations. These features are calculated from the instantaneous amplitude, instantaneous phase, and spectrum symmetry of the unknown signal. To recognize the modulations, an analysis of the features is carried out by means of a classifier based on Gaussian mixture models. The method was designed for 2ASK, 2FSK, MSK, BPSK, QPSK, and 16QAM modulations. They belong to the widely used digital modulations in recent communication systems. Simulation results demonstrating performance of the method are presented. The testing signals are corrupted by white Gaussian noise.

Keywords

Recognition of Digital Modulations, Classification of Digital Modulations, Gaussian Mixture Models

RIV year

2011

Released

01.04.2011

Publisher

Brno University of Technology

Location

Brno, Czech Republic

Pages from

15

Pages to

22

Pages count

8

URL

BibTex


@article{BUT74214,
  author="Anna {Kubánková} and Hicham {Atassi} and David {Kubánek}",
  title="Gaussian Mixture Models-based Recognition of Digital Modulations of Noisy Signals",
  annote="A novel approach to the recognition of digital modulations is presented. It is based on features that characterize a specific kind of modulation or a group of modulations. These features are calculated from the instantaneous amplitude, instantaneous phase, and spectrum symmetry of the unknown signal. To recognize the modulations, an analysis of the features is carried out by means of a classifier based on Gaussian mixture models. The method was designed for 2ASK, 2FSK, MSK, BPSK, QPSK, and 16QAM modulations. They belong to the widely used digital modulations in recent communication systems. Simulation results demonstrating performance of the method are presented. The testing signals are corrupted by white Gaussian noise.",
  address="Brno University of Technology",
  chapter="74214",
  institution="Brno University of Technology",
  number="1",
  volume="2",
  year="2011",
  month="april",
  pages="15--22",
  publisher="Brno University of Technology",
  type="journal article - other"
}