Specific Emitter Identification Based on Continuous Learning and Joint Feature Extraction br

Considering the problem of low recognition accuracy of Specific Emitter Identification (SEI) and highcost of single training, an SEI scheme based on incremental learning is proposed in this paper, multipleContinuous Incremental Deep Extreme Learning Machine(CIDELM) are designed. The Hilbert spectrumprojection and higher-order spectrum processed by Variational Mode Decomposition (VMD) are extracted fromthe original signal, and they are used as the Radio Fingerprint Feature (RFF) for classification afterdimensionality reduction. In the Extreme Learning Machine (ELM), the sparse self-encoding structure isintroduced to perform unsupervised training on multiple hidden layers, and the parameter search strategy isused to determine the best number of hidden layers and hidden nodes, realizing online multi-batch labeledsamples continuous matching. The results show that the algorithm can show good compatibility with differentmodulation modes, carrier frequencies and transmission distances, and can effectively identify multiple transmitters