Deep Hierarchical Network for Automatic Modulation Classification

Jinbo Nie,Shiyao Chen,Zunwen He,Shouliang Gong,Yan Zhang,Wancheng Zhang

doi:10.1109/access.2019.2928463

Jinbo Nie, Shiyao Chen + Show 4 more

Open Access

https://doi.org/10.1109/access.2019.2928463

Copy DOI

Abstract

In non-cooperative communication scenarios, automatic modulation classification (AMC) is the premise of information acquisition. It has been a difficult issue for decades due to the attenuation and interference during wireless transmission. In this paper, a novel deep hierarchical network (DHN) based on convolutional neural network (CNN) is proposed for the AMC. The model is designed to combine the shallow features with high-level features. Thus, it can simultaneously have global receptive field and location information through multi-level feature extraction and does not require any transformation of the raw data. To make full use of limited data, a new method is proposed to use signal-to-noise ratio (SNR) as a weight in training instead of working as an indicator of system robustness. Furthermore, some other deep learning methods have been used to explore whether they could improve the performance of the proposed model. Several new techniques have been chosen to be applied in the DHN at last. Then, a detailed analysis of the improvement in network performance is provided. Combination of the DHN and the weighted-loss can achieve more than 93% classification accuracy which is the best performance in an open source dataset.

Highlights

With the development of communication technologies, the modulation methods of signals have become rich and diverse
We propose a novel deep hierarchical network (DHN) based on the common convolutional neural network (CNN) network, and it does not need any data preprocessing
We propose a modulation classification network containing multi-scale feature, and optimize it with a new signal-to-noise ratio (SNR) loss

Summary

Introduction

With the development of communication technologies, the modulation methods of signals have become rich and diverse. INDEX TERMS Automatic modulation classification, deep hierarchical network, receptive field, signal-to-noise ratio weight. A simple convolutional neural network (CNN) [17] structure was the first deep learning method to solve the AMC problem.

Results

Conclusion