Research On Passive Localization Method Based On Sources Of Opportunity

With the improvement on marine technology and the increasing scarcity of land nature resources,human gradually turns to the ocean for resources.Marine noise pollution caused by day and night offshore drilling platforms,ocean going freighters and scientific sonar equipment bring a serious threat to the survival of marine life.Recently in the field of underwater acoustics,there has been a movement towards the use non-active distributed sound source to localize the target and extract environmental information.Ship source is a kind of sources of opportunity,which belongs to passive technology.The ship source is of the same high source level as active source and can be tracked using the Automatic Identification System(AIS).The thesis mainly focuses on the passive localization algorithm based on the Double-Correlation Function(DCF)using sources of opportunity.By improving the cross-correlation processing,we propose a localization algorithm based on sources of opportunity with higher localization accuracy and resolution.The DCF builds up the measurement replicas of the search area through sources of opportunity,and localize the unknown target source through matching the event data with library data.Therefore,the DCF could be understood as a generalized matching field processing,which is different from the conventional matching field processing in that the replicas are built according to environmental parameters.The DCF builds up the replicas according to measurement,which effectively avoids the degradation of localization performance caused by mismatch of environmental parameters.During the cross-correlation processing of DCF,the target will be localized by matching phase in the second cross-correlation processing,and the purpose of the first crosscorrelation processing is to extract the phase containing the sound source position information and eliminate the specificity of the sound source spectrum.Therefore,the accuracy of signal phase estimation directly affects the localization performance.According to Wiener-Khinchin theorem,the cross-correlation function in time domain and the cross-power spectrum in frequency domain is a couple of Fourier transform pairs.The theoretical basis of DCF in frequency domain will be proved by formula derivation.In the thesis,Generalized Cross-Correlation(GCC)will be introduced to process the cross-correlation vectors obtained from the first cross-correlation processing,and different GCC weighting function will be used to enhance the calculation weight of phase in the matching process to improve the localization accuracy.Secondly,the matching processing in the second cross-correlation processing is reformulated as solving a linear matrix equation,in which the correlation coefficient on each grid forms the unknown signal vectors to be solved.Due to the spatial sparsity of the vectors and the underdetermination of the system equation observed by two horizontally distributed sensors,the compressive DCF based on sparse reconstruction is proposed in the thesis.According to framework of Compressed Sensing(CS),the compressive DCF could achieve high resolution.In addition,the compressive DCF based on coherent model and incoherent model are proposed for distributed network,which could improve the localization accuracy and resolution through the spatial gain.Finally,an interpolation algorithm is implemented based on the Wave Guide Invariant(WGI),which tries to obtain the complete replicas of the search area by shifting the known cross-power spectrum on trajectory in frequency domain.It makes the localization method based on DCF using sources of opportunity more practical.The thesis not only verifies the feasibility and high-resolution characteristics of the localization algorithm in the model simulation,but also shows that the localization algorithm has certain theoretical and practical significance when applied to experiment in air condition and the SWell Ex-96 Experiment.In the end,the thesis summarizes the whole study work and prospects the future direction of passive localization based on sources of opportunity.