Research On Emitter Signal Sorting Techniques In Complicated Environments

In the current information age,radar signal sorting technology has become an important part of the electronic reconnaissance system,which has an important impact on the follow-up signal performance analysis.In the complex environment,the radar emitter signal sorting refers to the correct classification of the emitter signal from the same radar,due to the influence of many irrelevant signals in the complex environment,how to effectively sorting the emitter signals has become an urgent problem to be solved.In this paper,for the complex emitter signal environment,we first study the preprocessing of irrelevant signals which filter out unwanted signals and lay the foundation for post-signal sorting,then study single parameter PRI sorting and multi-parameter clustering.The work of this paper is as follows.1)Firstly,we study the pre-filtering of irrelevant signals in the environment,which can not affect the sorting of the emitter signals.In this paper,we study the notch filter based on Gauss Newton algorithm in time domain,and its convergence factor is improved to have better filtering characteristics.Because the time domain filter can not deal with the broadband signal,then,the linearly constrained minimum variance criterion is studied at the spatial filtering level,simulation results show that the time domain and space domain filtering algorithm can filter high-power irrelevant signals.2)After filtering,the pulse signal is parameterized for signal sorting.This paper focuses on the application of three clustering sorting models to multi-parameter emitter signal sorting for single-parameter sorting in cases where the pulse condition is complex or the number is more is not ideal.Because FCM clustering algorithm is more dependent on the initial clustering center,the simulation results show that the algorithm can be applied to signal sorting and the sorting effect is good in the case of high SNR.The clustering hyper-cubic box in FART clustering algorithm is studied and analyzed,in a certain learning rate and other parameters set,the same type of signal can be quickly and correctly assigned to the same hyper-cubic box,thus the signal sorting is completed.Finally,we propose to apply the algorithm of AP clustering which is used to affinity-propagation ideas to signal sorting,the AP algorithm uses the idea of treating each signal data as a cluster center for efficient sorting of signals in complex environments,comparing the performance of the three algorithms,it can be seen that the AP algorithm is more accurate and stable than the other two algorithms.3)After multi-parameter pre-sorting,this paper studies the single parameter signal sorting based on pulse repetition interval(PRI).The steps and principles of two kinds of histogram CDIF algorithm and SDIF algorithm are studied and analyzed,the two algorithms are compared,and the advantages and disadvantages are explained in detail.Because the histogram method can not remove the influence of sub-harmonics,this paper studies the improved PRI transform method.The simulation results show that the improved PRI transform method has a good effect on the sub-harmonics.In this paper,an improved sorting algorithm based on improved PRI transform and SDIF algorithm is proposed.The algorithm can avoid the sub-harmonics and can distinguish the azimuth pulse.The simulation shows that the sorting effect is good.