Automatic modulation recognition of DVB-S2X standard-specific with an APSK-based neural network classifier

• Published in: Measurement : journal of the International Measurement Confederation Vol. 151; p. 107257
• Main Authors: Ali, Ahmed. K., Erçelebi, Ergun
• Format: Journal Article
• Language: English
• Published: London Elsevier Ltd 01-02-2020; Elsevier Science Ltd
• Subjects: Algorithms; APSK; Automatic modulation recognition; Classifiers; Communications systems; DVB-S2X standard; Feature extraction; Machine learning; Neural networks; Pattern recognition; Radius-constellation; Random noise; Satellite communications standard; Signal classification; Signal processing; Software radio; APSK; DVB-S2X standard; Pattern recognition; Radius-constellation; Satellite communications standard
• ISSN: 0263-2241; 1873-412X
• DOI: 10.1016/j.measurement.2019.107257

•A simple DVB-S2X M-APSK classifier with accurate and robust performance.•Enhanced classification performance through new instantaneous feature set.•The combination of new features and the neural network provides robust performance.•Simulation results demonstrate the merits of new M-APSK recognizer against noise.
Modulation recognition of transmitted signals remains a significant concern in smart, modern communication systems such software-defined radios. Applying machine learning algorithms to interesting features extracted from the input signals are widely used for classification of such signals. Here, we propose leveraging novel higher order spectra features (HOSF) to a classification algorithm based on neural network properties to deal with modulation recognition problems specified for the M-APSK DVB-S2X modulation signals standard. The HOSF characteristics of signals are extracted under Additive white Gaussian noise (AWGN) channel, and four individual parameters are defined for distinguishing modulation signals from the set, {16, 32, 64} -APSK. This approach makes the recogniser more intelligent and improves its success rate of classification. The results illustrate excellent classification accuracy obtained at a low SNR of 0 dB, which demonstrates the potential of combining these proposed features with a neural network classifier for the application of M-APSK modulation classification.