Particle Swarm Optimization-Based Deep Neural Network for Digital Modulation Recognition

Wenzhe Shi,Yang Li,Dejun Liu,Yang Zhao,Xing Cheng

doi:10.1109/access.2019.2932266

Wenzhe Shi, Yang Li + Show 3 more

Open Access

https://doi.org/10.1109/access.2019.2932266

Copy DOI

Journal: IEEE Access	Publication Date: Jan 1, 2019
Citations: 35	License type: CC BY 4.0

Affiliation: China University of Petroleum, Beijing

Abstract

Modulation recognition is a major task in many wireless communication systems including cognitive radio and signal reconnaissance. The diversification of modulation schemes and the increased complexity of the channel environment put higher requirements on the correct identification of modulated signals. Deep learning (DL) is considered as a potential solution to solve these problems due to the superior big data processing and classification capabilities. This paper proposes an efficient digital modulation recognition method based on deep neural network (DNN) model. Furthermore, we present the particle swarm optimization (PSO) algorithm to optimize the number of hidden layer nodes of the DNN so as to solve the problem that the traditional DNN is trapped in local minimum values and the number of hidden layer nodes needs selecting manually. In this paper, we utilize the proposed PSO-DNN method to learn characteristics extracted from the modulated signal added by additive white Gaussian noise (AWGN) and to train the network, which can improve the performance of recognition under the condition of low signal-to-noise ratio (SNR). The experimental results demonstrate that the recognition rate on this algorithm has improved by 9.4% and 8.8% compared with methods that adopt conventional DNN and support vector machine (SVM) when SNR equals 0 and 1 dB, respectively. Besides, another experiment compared with the genetic algorithm (GA) also proves that our proposed algorithm is more effective in optimizing the DNN. The proposed method is easy to be implemented so that it has a broad development prospect in modulation recognition.

Highlights

In wireless communications, modulation recognition is a kind of technology which can realize smart reception, processing, and classification of modulated signals
We propose a novel method in the scenario of multiple modulation signal recognition in wireless communications, which applied the technique of signal preprocessing and the improved deep neural network (DNN) model
We introduce a novel particle swarm optimization (PSO) optimization scheme to improve the structure of DNN to obtain the global optimal number of hidden layer nodes, obtaining the optimal accuracy under the condition of low signalto-noise ratio (SNR) in a modulation recognition system

Summary

INTRODUCTION

Modulation recognition is a kind of technology which can realize smart reception, processing, and classification of modulated signals It plays an important role in intelligent control for civilian purpose and signals monitoring for military purpose under the scenario that the receiver does not know the modulation format the sender used [1], [2]. We propose a novel method in the scenario of multiple modulation signal recognition in wireless communications, which applied the technique of signal preprocessing and the improved DNN model. We introduce a novel PSO optimization scheme to improve the structure of DNN to obtain the global optimal number of hidden layer nodes, obtaining the optimal accuracy under the condition of low SNR in a modulation recognition system. The proposed PSO-DNN, the conventional DNN and the SVM method are tested versus SNR, followed by simulation results and comparative analyses.

DATA PREPROCESSING

FEATURE ENGINEERING

DEEP NEURAL NETWORK MODEL

MODULATION RECOGNITION BASED ON PSO-DNN

RECOGNITION PERFORMANCE

CONCLUSION