【摘要】 特定辐射源识别在民用频谱管理中起着重要作用。传统的深度神经网络方法在辐射源识别方面面临诸多挑战，包括训练时间长、能耗高以及计算稀疏性低。针对这些问题，设计了一种基于深度复数脉冲神经网络模型，该模型集成了脉冲神经层，并利用复数数据的固有特性增强信号表达能力，显著优化了计算效率并降低了硬件资源需求。测试结果表明，该模型的识别准确率达到了96%,单条数据的平均推理时间为0.19 ms,在模型参数规模、推理速度和推理数据能量消耗上均优于传统神经网络模型。更多还原

【Abstract】 Specific emitter identification(SEI) plays a crucial role in civilian spectrum management. Traditional deep neural network methods face many challenges in emitter identification, including extended training duration, high energy consumption, and low computational sparsity. To address these issues, a deep spiking complex neural network(S-CNet) model was designed, which integrates pulse neural layers and utilizes the intrinsic properties of complex data to enhance signal representation capabilities, significantly optimizing computational efficiency and reducing hardware resource requirements. The test results have shown that the recognition accuracy of this model reaches 96%, the average inference time for a single piece of data is 0.19 ms, and it is superior to the traditional neural network models in terms of model parameter scale, inference speed, and inference data energy consumption.更多还原