Automatic Modulation Classification Based on Time-Attention Mechanism and LSTM Neural Networks
Tongyao Liu, Shaowei Wu · 2024
The ongoing advancement of 5G technology, the rapid expansion of low-orbit satellite networks, and the extensive utilization of intelligent devices have led to a growing scarcity of electromagnetic spectrum resources. Intelligent communication technology and advanced spectrum management are crucial for enhancing spectrum utilization efficiency. As a key technology, the automatic classification and recognition of complex modulated signals have garnered significant attention, with deep learning-based methods becoming a research hotspot in recent years. This paper proposes an improved deep learning-based method for automatic modulation classification using a time-series LSTM-AL model (LSTM-attention-LSTM). The superiority of the proposed method is validated on the RML2016.10A and augmented RML2016.10A datasets. Experimental results show that the LSTM-AL method designed in this paper outperforms other prediction methods in terms of accuracy.