Passive Localization Methods For Intercepted Signals Of Uniform Linear Array Sonar

There are a lot of active sonar pulse signals such as continuous wave(CW),linear frequency modulation(LFM),hyperbolic frequency modulation(HFM)and broadband ship radiated noise signals in the marine battlefield environment.If these signals can be intercepted,the target recognition and localization will be realized through further signal processing.Combined with the characteristics of the underwater acoustic intercepted signals and new processing technologies of array signals,passive target localization methods have many advantages,such as better platform concealment,higher directional resolution,and farther detection distance.These methods are thus becoming an important research field in underwater acoustic signal processing.Among others,focused beamformer(FB),fractional fourier transform(FRFT)and compressed sensing(CS)have been extensively studied in passive target localization.However,some of the existing methods are suitable for narrow-band signal localization,some are suitable for localization in far-field environment,and some are only suitable for processing LFM signals.Therefore,the passive localization methods which are based on combination and improvement of the existing methods for multiple underwater acoustic intercepted signals are worth further study.The innovative research work in this dissertation is as follows:(1)Aiming at the passive localization of narrowband CW pulse signals in the near field environment,FB methods such as C-FB,MVDR-FB and MUSIC-FB were studied.A performance analysis method based on spatial spectrum pseudo-peak number estimation was proposed,and the performance of FB techniques applied to underwater target passive localization were analyzed from a new perspective.The pseudo-peak in the FB’s spectra was defined and the estimation method of pseudo-peak number was given.On this basis,the effects of signal-to-noise ratio(SNR),signal frequency,target distance and pseudo-peak detection threshold on the performance of the three methods were studied.Finally,a pseudo peak detection threshold setting and its classification method were proposed based on the convergence of pseudo-peak number,which provides technical support for the research of near-field passive localization methods in the following chapters of this dissertation.(2)The existing FRFT passive localization methods can achieve the target direction of arrival(DOA)estimation in the far field environment.If the existing methods are directly applied in the near field environment,there will be problems such as increasing DOA estimation error and inability to achieve passive ranging.A passive near-field wideband LFM direction finding and ranging method(FRFT-MUSIC-FB)based on FRFT and MUSIC-FB was proposed.The near-field spherical wave model for wideband LFM signal was established and the time-varying array manifold matrix corresponding to the LFM signals was transformed into a fixed array manifold matrix by using FRFT.Combined with the near-field FB technique and multiple signal characteristics(MUSIC)algorithm,the directions and distances of multiple broadband LFM signals will be estimated.The analysis results of simulation and measured data shown that the proposed method has many advantages,such as higher localization accuracy,stronger anti-noise and anti-multipath interference characteristics.(3)When the existing FRFT passive localization methods are used for HFM signals,the localization error will increase as the frequency of HFM signal is nonlinear.Using the piecewise linearization tactics,a method of DOA estimation for far-field broadband HFM signals(TPFRFT-MUSIC)and a method of passive direction finding and ranging for near-field broadband HFM signals(TPFRFT-MUSIC-FB)were proposed based on the time domain piecewise FRFT(TPFRFT)and MUSIC beamformer.The models of received wideband HFM signals were established for uniform linear array(ULA)sonar in the near or far field.The received signal was divided into multiple short-time signals and FRFT was applied to transform the traditional array data into new array data in short-time fractional fourier domain.At the same time,the time-varying array manifold matrix was transformed into multiple fixed array manifold matrixs.The spatial spectrum of each short-time signal was estimated by MUSIC method,and the spatial spectrum of HFM signal in the whole time was further obtained by summation.Finally,HFM signal localization was realized through spectral peak searching.The effectiveness of the methods were verified by simulation experiments.The influence of time domain piecewise number,SNR and sound source position on the performance of the methods were studied.Compared with the existing methods,the proposed methods have higher resolution and accuracy.(4)Aiming at the problems of the spatial mesh generation and complete array manifold construction for CS theory in underwater acoustic near-field target passive localization,a passive localization method(CS-FB)based on CS was proposed for near-field underwater acoustic signals.The distribution space of the signal sources was meshed according to the strategy of direction followed by the distance.For a certain direction in space,a complete array manifold matrix in this direction was constructed as a sensing matrix through uniform distance.A convex optimization model for the underwater target signal estimation was constructed based on the CS theory and the received signal model for ULA sonar with near-field sources.The passive direction finding and ranging of underwater targets were further realized.Simulation results shown that,the method can solve the problem of low range resolution of existing FB methods,and also can realize passive direction finding and ranging for multiple underwater targets.Compared with the existing methods,the proposed method has higher range resolution,smaller blind area and farther detection distance.(5)The resolution of the existing wideband beamformers will be reduced when they are applied to underwater acoustic wideband weak targets localization.A method of DOA estimation for underwater wideband weak targets based on coherent signal subspace(CSS)processing and CS theory was proposed.Under the CSS processing framework,wideband frequency focusing was accompanied by a two-sided correlation transformation.DOA estimation of underwater wideband targets was realized based on the spatial sparsity of the targets and the CS reconstruction algorithm.Through analysis and processing of simulation data and marine trial data,it was shown that this method can considerably improve the spatial spectrum of weak target signals,enhance the ability to detect them.Compared with the ISS-MVDR,CSS-MVDR and ISS-CS,this method has better performance of weak target detection and DOA estimation.