Radar Signal Sorting Algorithm Under The Complex Electromagnetic Environment

Radar signal sorting refers to the process of acquisition and classification treatment of mixed interleaved radar pulse signal.In the increasingly complex electronic warfare environment,radar signal sorting technology has gradually become the key link of the electronic reconnaissance system validity detection.At the same time,the quality of radar signal sorting properties has important influence on the late goals identification.However,in the face of increasingly complex space electromagnetic environment,radar signal parameters have become densely interleaved,and radar system with low probability interception appears constantly,the traditional methods based on radar signal sorting parameters have been unable to meet the requirements of electronic warfare system.Therefore,how to sort the radar emitter signal out from the complex electromagnetic environment is particularly urgent.Aiming at this problem,this paper carried out specific research.First,the radar emitter clustering methods are carried on the analysis,in all of the traditional radar signal sorting methods,fuzzy C means clustering classic(FCM)is widely used in the field of radar signal classification,the algorithm is flexible to classify radar signal samples,the clustering accuracy is high,but it is often necessary to preset a prior information,algorithm is easy to be dependent on the initial value and easy to fall into local optimum.Introduced the kernel function into the FCM algorithm,Fuzzy kernel clustering algorithm is based on the FCM,the sample data in the clustering method will be mapped to a nonlinear multi-dimensional space,the clustering algorithm is better than the FCM algorithm of the method,but still cannot get rid of the defect of depending on the initial value.Therefore,this paper put forward a fuzzy kernel clustering algorithm based on artificial fish swarm intelligent algorithm,artificial fish swarm algorithm parameter setting is simple,which has good global optimization effect,and does not depend on the initial value selection.Comparing the improved algorithm with the traditional FCM algorithm and the PSO-FCM clustering algorithm and the FSA-FCM clustering algorithm,the simulation results show that the separation of improved algorithm has higher accuracy,and improves the global searching ability of the algorithm and the separation speed to a certain extent.At present,the method of extracting in pulse characters have become the hot research topic by domestic and foreign scholars in the signal sorting,in the pulse characteristicsextracting methods,the time-frequency atoms in pulse extraction method are effective.However,the atom amount of time-frequency atoms is huge,the calculation of classical time-frequency atom method has a higher complexity,in order to improve the computational efficiency,this paper presents a improved method of extracting in pulse frequency atom,through analysis the advantages and principle of parameter search of the artificial fish swarm intelligent algorithm,the algorithm is used in pulse frequency atom extraction process improvement.For radar emitter signals with different modulation mode,according to the "class discrimination criterion" to extract intro-pulse features of the radiation source,the intercept characterization of atoms of the low probability radar emitter signals can be obtained.The simulation results show that,the improved classification method has good clustering effect for radar signals with different modulation when the SNR is relatively low.