Specific Emitter Identification Based On Spectral Analysis

Specific emitter identification plays an important role in electronic reconnaissance.It’s also one of the core technologies to identify threat targets in electronic countermeasures.However,as the electromagnetic environment becomes more and more complex,identification of unknown emitters has become the technical bottleneck of effective electronic reconnaissance.In order to solve the problem of the identification of known and unknown emitters in combat space,the discrimination of unknown emitters,known emitters and unknown radiation sources are studied.Firstly,the causes of the unintentional modulation characteristics in the pulse of the emitter are analyzed.The working principle of the emitter transmitter are conbined.The factors of establishing the emitter model are grasped.The mathematical model of the emitter are constructed.Secondly,combined with time-frequency transform principle,this paper proposes to use the particle swarm optimization algorithm to optimize the extraction parameters of time-frequency ridge.The time-frequency distribution region of intentionally modulated signals is determined and suppressed.The signal is reconstructed by using time-frequency inversion.Thirdly,based on the idea of spectral analysis,combined with bisspectral,cyclic spectral and attention mechanism,a lightweight binary channels neural network is designed.This paper proposes the emitter identification algorithm based on spectral analysis and attention mechanism,which can adaptive extract and fuse the features,then realize the identification of known emitters.In addition,based on the metric distance,the relationship between the individual and the metric distance of the radiation source is fused.The neural network based on spectral analysis and attention mechanism is transferred as the feature extractor.This paper proposes the discrimination algorithm based on the metric distance to discriminate the unknown emitters.Finally,for the discriminated known emitters,the emitter identification algorithm based on spectral analysis and attention mechanism is used to complete the identification.For the discriminated unknown emitters,this paper proposes the sampling square integrated bispectra and the improved mesh density clustering algorithm combined with metric distance,which can solve the unknown emitter identification problem.At the same time,this paper gives out the scheme of emitter identification system.The simulation results show that,under the condition of signal-to-noise ratio of 0d B,the discrimination accuracy of the known and unknown emitter reaches 90.37%.The recognition rate of the known emitter and the unknown emitter can reach 91.13% and 81.00% respectively.Bisedes,under the condition of signal-to-noise ratio of 0d B,those three accuracy indexes can reach 91.70%,94.50% and 84.80% respectively,which verifies the rationality of the system and provides a new idea for the identification of the emitters.