基于多参数时间序列相关性的雷达脉冲分选方法OA

Aeronautical electronic reconnaissance equipment can intercept,record and analyze the electromagnetic signals from radar emitters,and has the advantages of high mobility and wide surveillance coverage.As a critical type of electronic information equipment,the accurate extraction,deinterleaving,and identification of radar emitter signals are a key step in electronic reconnaissance.However,the complex electromagnetic environment has dense pulse streams and wide signal coverage,and practical scenarios often involve complex PRI modulation and pulse loss.Under such non-ideal conditions,traditional signal deinterleaving algorithms suffer from dependence on hyperparameters and limited deinterleaving accuracy.Therefore,an intelligent radar signal deinterleaving method based on multi-parameter time-series correlations is proposed.The method utilized time of arrival(TOA),radio frequency(RF),and pulse width(PW)parameters,employing a recurrent neural network to learn the long-term temporal patterns in the pulse stream.Experiment results showed that the proposed method could obtain robust performance under non-ideal conditions,which can achieve an accuracy of over 90％even if the pulse loss ratio reached 70％.