A Method for Cross-receiver Specific Emitter Identification Based on CBAM-CNN-BDA

To solve the problem that recognition accuracy drops due to the receiver is replaced in the application of specific emitter identification (SEI) technology, a method for cross-receiver SEI based on CBAM-CNN-BDA is proposed. In this paper, transfer learning is used to match feature distributions, thereby reducing the influence of cross-receiver factors on recognition performance. Particularly, the Convolutional Block Attention Module (CBMA) is added into the Convolutional Neural Network (CNN) in order to extract deep features. Then, we map the features of different distributions to a high-dimensional Reproducing Kernel Hilbert Space (RKHS) with the method of Balanced Distribution Adaptive (BDA). Finally, K-Nearest Neighbor (KNN) is used to identify specific emitters and unsupervised transfer learning between different receivers is realized. The proposed framework in this paper has a prefect performance on the modeling simulation signals of cross-receiver SEI. Simultaneously, the experiment with real signals which are collected from Automatic Identification System (AIS) proves that CBAM-CNN-BDA can effectively improve the identification accuracy and outperform than the traditional algorithm.