Data Augmentation for Deep Learning-Based Radio Modulation Classification

Huang, Liang; Pan, Weijian; Zhang, You; Qian, Li Ping; Gao, Nan; Wu, Yuan

doi:10.1109/access.2019.2960775

Cited by 132 publications

(62 citation statements)

References 41 publications

(60 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Data augmentation techniques are considered as a proper solution for the drawbacks of insufficient training data. In [102], for example, a state-of-the-art LSTM-based AMR was presented to evaluate Gaussian noise, rotation, and flip methods of radio signals augmentation in both training and inference phases. Similarly in [103], RNN and LSTM were investigated for AMR.…”

Section: ) Rnn-based Methodsmentioning

confidence: 99%

Machine Learning Based Automatic Modulation Recognition for Wireless Communications: A Comprehensive Survey

et al. 2021

View full text Add to dashboard Cite

The rapid development of information and wireless communication technologies together with the large increase in the number of end-users have made the radio spectrum more crowded than ever. Besides, providing a stable and reliable service is challenging, as electromagnetic environments are evolving and becoming more sophisticated. Accordingly, there is an urgent need for more reliable and intelligent communication systems that can improve the spectrum efficiency and the quality of service to provide agile management of network resources, so as to better meet the needs of future wireless users. Specifically, Automatic Modulation Recognition (AMR) plays an essential role in most intelligent communication systems especially with the emergence of Software Defined Radio (SDR). AMR is an indispensable task while performing spectrum sensing in Cognitive Radio (CR). Thanks to the significant advancements in Deep Learning (DL) applications, new and powerful tools have been provided which can tackle problems in this space. Thus, today, integrating DL models into AMR has gained the attention of many researchers. This work aims to provide a comprehensive state-of-the-art review of the most recent Machine Learning (ML) based AMR methods for Single-Input Single-Output (SISO) and Multiple-Input Multiple-Output (MIMO) systems. Furthermore, the architecture of each model will be identified along with a detailed comparison in terms of specifications and performance. Finally, an outline of the open problems, challenges, and potential research directions is provided along with discussion and conclusion.

show abstract

Section: ) Rnn-based Methodsmentioning

confidence: 99%

Machine Learning Based Automatic Modulation Recognition for Wireless Communications: A Comprehensive Survey

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Compared to random cropping, random erasing does not impact data integrity. In the field of AMC, Weijian Pan et al [12] effectively expanded training data by rotating constellation diagram and adding noise to it. Shiyao Chen et al [13] used generative adversarial network (GAN) [14] for data augmentation.…”

Section: ) Data Augmentation Methodsmentioning

confidence: 99%

“…The first part is signal embedding module. The data format in RML2016.10a is 2 × 128 , which can be directly used as input to LSTM, so the existing work such as [7], [12], [13] is to input IQ data directly into LSTM for processing. If a set of data is viewed as a sentence, then for the LSTM, the sentence is made up of 128 words, each of which is a vector of length 2.…”

Section: Single-layer Lstm Model Based On Attention Mechanismmentioning

confidence: 99%

Automatic Modulation Classification Scheme Based on LSTM With Random Erasing and Attention Mechanism

et al. 2020

View full text Add to dashboard Cite

Automatic modulation classification (AMC) is a key technology of cognitive radio used in non-cooperative communication. Recently, deep learning has been applied to AMC tasks. In this paper, an AMC scheme based on deep learning is proposed, which combines random erasing and attention mechanism to achieve high classification accuracy. Firstly, we propose two data augmentation methods, random erasing at sample level and random erasing at amplitude/phase (AP) channel level. The former replaces training samples with noise information, while the latter replaces AP channel information of training samples with noise information. Erased data segments are randomly stitched to enable training data expansion. Training data of different qualities enables deep learning model to have stronger generalization capability and higher robustness. Then, we propose a single-layer Long Short-Term Memory (LSTM) model based on attention mechanism. In the first part of this model, we propose the signal embedding, which enables the input to contain modulation information more comprehensively and accurately. Then hidden state output by LSTM is input into the attention module, and weighting is applied to the hidden state to help the LSTM model capture the temporal features of modulated signals. Compared to a model without attention mechanism, this model has faster convergence speed and better classification performance. Lastly, we propose a random erasing-based test time augmentation (RE-TTA) method. Test data is randomly erased for multiple times and classification results are comprehensively evaluated, in order to further improve classification accuracy. Experimental results on dataset RML2016.10a show that classification accuracy of the proposed scheme is competitive compared with the state-of-the-art methods.

show abstract

“…Huang et al applied a few data augmentation methods, such as rotation, flip, and Gaussian noise, on the RadioML Dataset. Their experiments showed that the rotation method yielded the best accuracy [32], and needed fewer data to achieve relatively good results. For the convenience of application, exploring lightweight networks is also very important for AMC problems.…”

Section: B Deep Neural Network Methodsmentioning

confidence: 99%

Automatic Modulation Classification Based on Hierarchical Recurrent Neural Networks With Grouped Auxiliary Memory

Zang

2020

IEEE Access

View full text Add to dashboard Cite

As a valuable topic in wireless communication systems, automatic modulation classification has been studied for many years. In recent years, recurrent neural networks (RNNs), such as long shortterm memory (LSTM), have been used in this area and have achieved good results. However, these models often suffer from the vanishing gradient problem when the temporal depth and spatial depth increases, which diminishes the ability to latch long-term memories. In this paper, we propose a new hierarchical RNN architecture with grouped auxiliary memory to better capture long-term dependencies. The proposed model is compared with LSTM and gated recurrent unit (GRU) on the RadioML 2016.10a dataset, which is widely used as a benchmark in modulation classification. The results show that the proposed network yields a higher average classification accuracy under varying signal-to-noise ratio (SNR) conditions ranging from 0 dB to 20 dB, even with much fewer parameters. The performance superiority is also confirmed using a dataset with variable lengths of signals. INDEX TERMS Automatic modulation classification (AMC), recurrent neural networks (RNNs), hierarchical recurrent structure, long-term memory

show abstract

Data Augmentation for Deep Learning-Based Radio Modulation Classification

Cited by 132 publications

References 41 publications

Machine Learning Based Automatic Modulation Recognition for Wireless Communications: A Comprehensive Survey

Machine Learning Based Automatic Modulation Recognition for Wireless Communications: A Comprehensive Survey

Automatic Modulation Classification Scheme Based on LSTM With Random Erasing and Attention Mechanism

Automatic Modulation Classification Based on Hierarchical Recurrent Neural Networks With Grouped Auxiliary Memory

Contact Info

Product

Resources

About