Modulation Type Identification And Parameter Estimation Of Typical Radar Signals Under Low Signal-to-noise Ratio

With the complexity of modern electromagnetic environment,radar system plays an increasingly important role in the fields of communication,navigation and electronic countermeasure.Compared with traditional radar systems,some modern new radar systems differ greatly in radar transmission waveform,number of antennas,antenna placement,signal processing method and so on.The anti-interception capability of modern new radar system has been significantly improved,which undoubtedly brings great challenges to electronic reconnaissance technology.In order to distinguish the enemy radar signal and accurately estimate the radar signal related parameters,not only the traditional pulse modulation parameters need to be measured,but also the in-pulse modulation characteristics of the radar signal need to be analyzed and distinguished.In this paper,the modulation type recognition methods of typical radar signals under low signal-to-noise ratio are studied,and a method based on the fourth-order origin moment of fractional spectrum is proposed.The feasibility of the method is verified by computer simulation.Then,the parameter estimation of linear frequency modulation signals under low signal-to-noise ratio is studied.This paper will introduce the main research work around the following aspects:First,the basic theory of fractional Fourier transform is analyzed,the definition and properties of fractional Fourier transform are introduced,the advantages of fractional Fourier transform in the field of non-stationary signal processing are shown through simulation experiments,and a sampling fast fractional Fourier transform algorithm based on discrete scaling method is studied.Secondly,the signal models of three typical radar signals are introduced,and the fractional Fourier amplitude maps of corresponding signals are given.Then,the identification method of typical radar signal modulation types based on fractional Fourier transform is introduced.For the poor performance of traditional radar signal modulation type methods in low signal-to-noise ratio,a method based on fourth-order origin moment of fractional Fourier spectrum is presented,which is a coarse-to-fine method.This method first roughly identifies the typical radar signal,and then distinguishes it into two types: the modulation signal and the frequency modulation signal.Then,it finely identifies the signal within the class,so as to realize the modulation type recognition of the typical radar signal.Simulations under different SNR conditions and comparison and analysis of the recognition probability results verify the anti-noise performance of the proposed recognition method under low SNR.Thirdly,aiming at the parameter estimation of single-component linear frequency modulation signal under low signal-to-noise ratio,a normalized parameter estimation algorithm based on fourth-order origin moment of fractional order spectrum is presented.Through the good anti-noise performance of fourth-order origin moment of fractional order Fourier,the optimal transformation order can be better determined at low signal-to-noise ratio,thus improving the anti-noise ability of parameter estimation algorithm,and then combining the discrete fast fractional Fourier transform algorithm.Signal parameters are calculated by peak search.The comparison and analysis of simulation experiments verify that the algorithm guarantees the accuracy of signal parameter estimation and has good anti-noise performance.Finally,for parameter estimation of multi-component linear frequency modulation signals under low signal-to-noise ratio,a parameter estimation normalization method based on fourth-order origin moments of fractional frequency spectrum is presented,which combines CLEAN idea.By separating the signal components within the multi-component linear frequency modulation signal,each signal component is estimated one by one until no valid spectral peak is searched.The parameter estimates of all signal components of a multicomponent linear frequency modulation signal can be obtained.The parameter estimation and anti-noise performance of the algorithm at low signal-to-noise ratio are verified by simulation.