Application and Optimization of Deep Learning in Semantic Recognition of Communication Signals

Semantic recognition of communication signals is very important for ensuring communication quality, improving efficiency and maintaining national security. Deep learning, especially Convolutional Neural Network (CNN) and Long-term Memory Network (LSTM), has shown great potential in the field of comm...

Full description

Saved in:
Bibliographic Details
Published in2025 International Conference on Electrical Drives, Power Electronics & Engineering (EDPEE) pp. 1165 - 1169
Main Authors Meng, Hao, Lei, Yingke, Liu, Changming, Jin, Hu, Wang, Jin, Teng, Fei
Format Conference Proceeding
LanguageEnglish
Published IEEE 26.03.2025
Subjects
Accuracy
Attention mechanism
Attention mechanisms
Attention-CNN-LSTM
Communication signals
Computational modeling
Convolutional neural networks
Data models
Deep learning
Feature extraction
Long short term memory
Optimization
Semantic recognition
Semantics
Online AccessGet full text

Cover

More Information
Summary:Semantic recognition of communication signals is very important for ensuring communication quality, improving efficiency and maintaining national security. Deep learning, especially Convolutional Neural Network (CNN) and Long-term Memory Network (LSTM), has shown great potential in the field of communication signal recognition because of its powerful feature extraction and sequence modeling capabilities. In this study, a model combining 1D CNN and LSTM is proposed, and attention mechanism is introduced to enhance the model's ability to capture key features in the signal. The experimental results show that the model with attention mechanism is superior to the traditional model in recognition accuracy and processing efficiency. Experiments on IEEE 802.11 WLAN signal data set show that the model is effective, and the recognition accuracy of Attention-CNN-LSTM model reaches 95.2%, which is 3.1 percentage points higher than that of the model without attention mechanism. In addition, the model processing time has also been significantly improved.
DOI:10.1109/EDPEE65754.2025.00210