Radar Signal Ambiguity Function Main Ridge Section Geometry Feature Extraction Method

Modern warfare under the condition of informatization is inseparable from timely,accurate and reliable radar countermeasure intelligence.Radar countermeasure in radar electronic warfare is the main means to obtain radar countermeasure intelligence.However,with the increasing complexity of the electromagnetic environment,the density index of radar radiation source signals surges,which puts forward higher requirements on the effectiveness,reliability and timeliness of the signal processing algorithm.Traditional radar emitter signal sorting is the main characteristic parameters of the Radio Frequency(RF),Pulse Width(PW),Pulse Amplitude(PA),Time of Arrival(TOA),Direction of Arrival(DOA)five classic parameters,but due to the complex modulation mode diversification,the modern radar parameters,change characteristics of fast,only using conventional parameter is difficult to achieve ideal separation identification accuracy.Since the ambiguity function(AF)can reflect the intrinsic information of the signal in detail,this paper proposes a feature extraction method based on the AF to further improve the accuracy of radar signal sorting and recognition.Firstly,a method,based on ambiguity function main ridge(AFMR)is presented.This method extracts the main ridge slice from received signals,via a genetic algorithm algorithm.Then,the energy moment and fuzzy entropy are extracted as the feature vector which can describe a signal's fuzzy energy concentration and complexity.The simulation experiment of this method shows that the proposed feature fully reflects the distribution characteristics of fuzzy energy of AFMR.The success rate of the proposed feature in sorting six types of typical radar signals under the condition of greater than 0dB remains above 94.33%.Even with a signal-to-noise ratio of-2db,the success rate of sorting can still reach 89.17%.Under the condition of 0-20 db dynamic signal-to-noise ratio,the average sorting success rate is 99.55%.Theoretical analysis and sorting time comparison experiment also show that the proposed method has low algorithm complexity,which proves its effectiveness and timeliness.Then,a morphological feature extraction method of the AFMR based on polar coordinate transformation was proposed.The main ridge of AF was transformed into polar coordinate domain to form a closed geometric image.the average area,density and roundness of the image were extracted as the feature vectors of signal sorting.And Simulation experiments show that the proposed features fully reflect the changing characteristics and complexity characteristics of AFMR.Under the condition that SNR is no less than 0dB,the proposed features can effectively sort out various radar signals with a high accuracy.It also has high accuracy under the condition of dynamic SNR.Compared with other methods,it has higher accuracy and lower algorithm complexity.It not only has good class cohesion,but also has strong anti-interference ability,which can be used as an effective supplement to classical five-parameter sorting features.