Dual-Channel Hybrid Neural Network for Modulation Recognition

Min Lu,Tianjun Peng,Guangxue Yue,Xiangbai Liao,Bolin Ma

doi:10.1109/access.2021.3081617

Abstract

In the blind recognition of wireless communication modulation based on deep learning, how to further improve the recognition accuracy has become a key research issue. In this work, we propose a dual-channel hybrid model termed as CLDR (convolutional long short-term deep neural and residual network). The CLDR consists of the convolutional long short-term deep neural network (CLDNN) and the residual network (ResNet), where CLDNN is to reduce the variations in the spectrum and time, ResNet is to avoid gradient vanishing or exploding. In addition, we design an exponential curve decay adaptive cyclical learning rate method to decrease the training time cost of the neural network model. This method eliminates the need to experimentally search the optimal learning rate as with the fixed learning rate policy. It also avoids the slow convergence of the model due to the excessive attenuation amplitude as with the triangular learning rate policy. We test the feasibility of the CLDR model and discuss the influence of the exponential decay cyclical learning rate on the training of CLDR model based on the RadioML2016.10b public dataset. Simulation results show that the CLDR model yields a recognition accuracy of 93.1% at high SNRs, which effectively reduces the influence of external environment such as noise and fading on recognition accuracy. The training time cost of CLDR using exponential cyclical learning rate is reduced by 14.6% and 32.1% compared with triangular and fixed methods. Therefore, the exponential cyclical learning rate policy effectively reduces the training time cost of the model in the same classification accuracy.

Highlights

With the rapid development of technology such as pattern recognition and signal processing [1], using deep learning to classify communication modulated signals has gradually become the focus of research
The results show that the training accuracy and the validation accuracy of the CLDR model is the highest compared with the rest models under three different learning rate policies, and the CLDR model runs 18, 26, and 24 epochs to achieve 60% training accuracy under the fixed learning rate policy, triangular cyclical learning rate policy, and exponential cyclical learning rate policy. the CLDR model runs fewer training epochs to reach the training accuracy of other models
We show the recognition effect of the convolutional long short-term deep neural network (CLDNN) network model combined with different network models under the fixed learning rate policy, as shown in Figure 9, the results show that the recognition performance of our CLDR model is significantly better than other combination models

Summary

Introduction

With the rapid development of technology such as pattern recognition and signal processing [1], using deep learning to classify communication modulated signals has gradually become the focus of research. According to the training history, the propose CLDR achieves the least loss rate and the fastest convergence rate compared with other models under different learning rate policies.

Results

Conclusion