基于空时编码的雷达脉冲辐射源识别算法

Radar Pulse Radiation Source Identification Based on Spatiotemporal Encoding

随着宽带电子侦察接收机的广泛使用,同时进入接收机的信号越来越多,输出的数据量越来越大,传统的雷达辐射源识别方法很难适应分析处理的实时性和准确性要求。为解决这一问题,提出基于空时编码的雷达脉冲辐射源识别算法。通过设计一种空时编码算法来对雷达辐射源全脉冲数据中的载频、脉宽及重频等参数进行编码和特征提取,形成包含全脉冲数据时空信息的特征序列。这些特征序列既保留了雷达全脉冲数据的主要信息,又实现了对雷达全脉冲数据的简化,为利用深度学习和大数据处理技术实现对雷达辐射源快速精准识别奠定基础。利用长短时记忆网络(Long Short-Term Memory,LSTM)深度神经网络模型,结合提取的空时特征序列,通过大数据学习训练,实现了对7型雷达辐射源的精确识别。在数据集上的对比实验表明,提出的识别方法在处理识别精度和普适性上优于目前的主流算法,具有较强的鲁棒性。

With the use of wideband electronic reconnaissance receiver,more and more signals are entering the receiver,and the output data is becoming more and more complex.The traditional radar radiation source identification method is difficult to meet the accuracy requirements of analysis and processing.To solve this problem,this paper proposes a radar pulse source recognition algorithm based on spatiotemporal encoding.This paper designs a spatiotemporal encoding algorithm to code and extract the RF,PW and PRI features in radar radiation source full pulse data and form the feature sequence containing the spatiotemporal information of the full pulse data.These feature sequences not only retain the main information of radar full pulse data,but also simplify the radar full pulse data,which lays the foundation for the rapid and accurate recognition of radar radiation sources by using deep learning and big data processing technology.Then,the model of Long Short-Term Memory(LSTM)deep neural network is used to recognize the radiation source of 7 radar emitters by using the spatiotemporal feature sequence and big data learning strategy.The comparison experiments on the dataset show that the proposed method is superior to the current mainstream algorithm in term of recognition accuracy and universality with has strong robustness.