Feature Extraction, Modulation and Recognition of Mixed Signal Based on SVM

TL;DR

This paper presents methods for recognizing modulation types of mixed signals using feature extraction techniques and SVM classification, demonstrating effective differentiation among various digital modulation schemes.

Contribution

It introduces a combination of likelihood-based and feature-based modulation recognition methods, utilizing multiple signal features for improved classification accuracy.

Findings

Effective differentiation of six digital modulation signals.

Use of multiple features enhances recognition accuracy.

Demonstrated applicability to mixed signals in simulation.

Abstract

This paper introduces likelihood-based and feature-based modulation recognition methods. In the feature-based modulation simulation part, instantaneous feature, cyclic spectrum, high-order cumulants, and wavelet transform features are used as the entry point, and six digital signals including 2ASK, 4ASK, BPSK, QPSK, 2FSK and 4FSK are simulated, showing the difference of signals in multiple dimensions