基于加伯变换和希尔伯特变换的信号调制识别方法
By combining the feature matrices of Garber transform and Hilbert transform with IQ-path data, and integrating deep learning networks, especially attention mechanisms, the performance bottleneck of traditional FSK signal recognition in complex environments has been solved, achieving efficient and accurate signal recognition.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN JIUZHOU SOFTWARE CO LTD
- Filing Date
- 2025-06-10
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional FSK signal recognition methods are limited in time-frequency resolution under complex electromagnetic environments, feature extraction relies on experience, and have poor performance with low signal-to-noise ratio, making it difficult to adapt to the variability of FSK. Existing technologies do not fully utilize the synergistic advantages of Garber transform and Hilbert transform.
By combining the feature matrices extracted by the Discrete Garber Transform and Hilbert Transform with IQ-path data as input, signal recognition is performed through a deep learning network. In particular, an attention mechanism is introduced to improve feature fusion and noise resistance.
It significantly improves the recognition accuracy and robustness of FSK signals in complex environments, and enhances the adaptability and generalization ability to different modulation parameters.
Smart Images
Figure CN120415974B_ABST