Signal reconstruction contrastive learning method utilizing window features

With the rapid growth of signal data in the industrial internet, the issue of missing labels has become increasingly prominent, making self-supervised learning a critical solution. To address the problems of coarse feature granularity, unstable representation, and weak transferability in existing contrastive learning methods for signal recognition tasks, a window-based signal reconstruction contrastive learning approach was proposed. The method divides feature maps into multiple fixed windows and introduces local similarity constraints to construct a fine-grained contrastive structure. It also incorporates a signal reconstruction module to enhance the stability and semantic consistency of feature representations. Furthermore, a loss function integrating reconstruction error was designed to improve the feature fitting capability to the original signals. Experiments on three signal recognition datasets—RML, ADS-B, and CSI—show that the proposed method achieves up to 28.32% higher performance compared to other contrastive approaches. Cross-dataset transfer accuracies reach 65.36%, 66.17%, and 66.96%, respectively, significantly outperforming existing methods and demonstrating strong transfer and generalization capabilities.

To address the problems of coarse feature granularity, unstable representation, and weak transferability in existing contrastive learning methods for signal recognition tasks under missing label scenarios.

A window-based signal reconstruction contrastive learning approach was proposed. The method divided feature maps into multiple fixed windows and introduced local similarity constraints to construct a fine-grained contrastive structure. It incorporated a signal reconstruction module to enhance representation stability and semantic consistency, and designed a loss function integrating reconstruction error to improve feature fitting capability.

Experiments on three signal recognition datasets (RML, ADS-B, and CSI) show the proposed method achieves up to 28.32% higher performance compared to other contrastive approaches. Cross-dataset transfer accuracies reach 65.36%, 66.17%, and 66.96% respectively.

The proposed method significantly outperforms existing methods in both recognition performance and transfer learning capabilities, demonstrating strong generalization ability for industrial signal recognition tasks with missing labels.