Enhancing Automatic Modulation Recognition for IoT Applications Using Transformers

Automatic modulation recognition (AMR) is critical for determining the modulation type of incoming signals. Integrating advanced deep learning approaches enables rapid processing and minimal resource usage, essential for IoT applications. We have introduced a novel method using Transformer networks for efficient AMR, designed specifically to address the constraints on model size prevalent in IoT environments. Our extensive experiments reveal that our proposed method outperformed advanced deep learning techniques, achieving the highest recognition accuracy.