Towards recurrent neural network with multi-path feature fusion for signal modulation recognition

Lei, Zihang; Jiang, Mengxi; Yang, Guangsong; Tianmin, Guan; Huang, Peng; Gu, Yu; Xu, Zhenghua; Ye, Qiubo

doi:10.1007/s11276-021-02877-8

Cited by 8 publications

(3 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to test the effectiveness of Intra-InterNet network under complex noise environment, seven modulated signals (8PSK, BPSK, GFSK, PAM4, 16QAM, 64QAM and QPSK) in Tim O 'Sheatim's RadioML2016.10a data set [30][31][32] were selected as data samples in this paper. The dataset is generated using the open-source software radio platform GNU Radio.…”

Section: Simulation Datamentioning

confidence: 99%

An improved CNN‐LSTM network for modulation identification relying on periodic features of signal

Zhou,

Li,

Wang

2023

IET Communications

View full text Add to dashboard Cite

Aiming at the problem that modulated signals' classification and recognition accuracy is not high in the complex and changeable electromagnetic environment, this paper improves the CNN‐LSTM network and proposes a modulation recognition method of the CNN‐LSTM network (Intra‐InterNet) based on the periodic characteristics of the signal. The method uses the feature extraction ability of the convolutional neural network to extract the correlation features within the cycle and between adjacent cycles of different kinds of modulated signals. At the same time, the relevant features of the signals of non‐adjacent cycles are extracted by using the characteristics of the LSTM. This feature extraction method effectively enhances the feature extraction capability of the network by utilizing the periodicity of the modulated signal. Through the identification and verification of various modulation signals, the experimental results show that: under the signal‐to‐noise ratio of −6 dB ∼ 6 dB, the average identification accuracy rate reaches 87.29%, which effectively improves the accuracy of signal modulation mode identification.

show abstract

Section: Simulation Datamentioning

confidence: 99%

An improved CNN‐LSTM network for modulation identification relying on periodic features of signal

Zhou,

Li,

Wang

2023

IET Communications

View full text Add to dashboard Cite

show abstract

“…To more intuitively demonstrate the performance of the network proposed in this paper, five types of networks, ACGAN [24], MPFNet [14], MCLDNN [15], CLDNN [25], and ResNet [26], are selected for comparing with TC-MSNet. The structural details of each network are shown in Table 7.…”

Section: The Performance Comparison Experimentsmentioning

confidence: 99%

“…Those works also verified the importance of spatial features in AMR. The approaches in [14,15] achieved the extraction and fusion of multi-scale features by optimizing the network structures, and the recognition results showed that Multi-Scale Spatial (MSS) features can effectively advance the performance of neural networks for AMR.…”

Section: Introductionmentioning

confidence: 99%

TC-MSNet: A Multi-feature Extraction Neural Framework for Automatic Modulation Recognition

Wang,

Feng,

Song

et al. 2023

Preprint

View full text Add to dashboard Cite

In the approaches of Automatic Modulation Recognition (AMR), modulation modes with similar characteristics are prone to be confused by the adverse factors such as noise, inevitably bringing challenges to the accuracy of recognition. Aiming at this kind of problems, this paper proposes TC-MSNet which is a novel multi-scale spatial-temporal features collaboration neural network based on deep learning. TC-MSNet extracts the Temporal Correlation (TC) features and the Multi-Scale Spatial (MSS) features respectively for enhancing the diversity of features extraction, meanwhile, a fusion strategy constructed with convolu-tional attention mechanism implement sufficiently collaboration of multi-scale features. Additionally, the Bilinear-Pooling Mechanism (BPM) is adopted to capture the differentiation of fine-grained features, eliminating the confusion among distinct modulation modes to promote AMR accuracy. The simulation experimental results show that, on the RML2016.10a and RML2016.10b, the proposed TC-MSNet not only has better performance compared with existing deep learning-based AMR methods but also possesses stronger ability to eliminate confusion of similar features.

show abstract

Computer vision-driven forest wildfire and smoke recognition via IoT drone cameras

Wang,

et al. 2024

Wireless Netw

View full text Add to dashboard Cite

Towards recurrent neural network with multi-path feature fusion for signal modulation recognition

Cited by 8 publications

References 35 publications

An improved CNN‐LSTM network for modulation identification relying on periodic features of signal

An improved CNN‐LSTM network for modulation identification relying on periodic features of signal

TC-MSNet: A Multi-feature Extraction Neural Framework for Automatic Modulation Recognition

Computer vision-driven forest wildfire and smoke recognition via IoT drone cameras

Contact Info

Product

Resources

About