Research On Joint DOA And Frequency Estimation Based On Spatial-Temporal Sparse Design

Direction of arrival(DOA)estimation is an important research topic in array signal processing.In practical applications,a uniform linear array(ULA)with the same spacing between adjacent antennas is generally used.When the number of antennas is limited,the expected array aperture and resolution may not be achievable using the ULA,therefore reachers turn to study sparse arrays.For a given number of achievable antennas,sparse array may have larger array aperture and higher angular estimation resolution than the ULA.Similarly,the use of sparse sampling in the time domain may improve the frequency estimation resolution.This thesis considers the optimal design of spatial sparse array and temporal sparse sampling around the joint estimation problem of DOA and Doppler frequency under the condition of limited physical resources.The main contribution and novelty are summarized as follows:Firstly,we propose a spatial-temporal nested sampling method with identical temporal parameters,using nested arrays in the spatial domain and nested sampling in the temporal domain,and restricting all arrays select the same time-domain nesting parameters.Such sparse sampling method can generate a larger virtual aperture in both the space and time domain,and hence increase the degree of freedom and the estimation resolution.Based on the virtual aperture,an equivalent received signal model is developed.In the model,the source power repleces the data sequence.Since the source power assumes constant values,the problems becomes joint DOA and Doppler frequency estimation of coherent signals.we consider to employ the spatial smoothing technique to construct the covariance matrix,and then use the MUSIC algorithm for joint estimation.In addition,when the maximum achievable number of physical antennas in space and sampling points in time are limited,the sparse sampling method is compared with spatial-temporal coprime sampling method with identical temporal parameters and spatial-temporal uniform sampling method with identical temporal parameters in terms of the degrees of freedom and the maximum number of estimable sources,which illustrates the advantages of the proposed sparse sampling method.Secondly,limited to the effect of antenna coupling in practical engineering applications,the first-stage subarray of the nested array does not have a sparse structure,hence the coupling effect is relatively large,therefore,this thesis uses coprime arrays in the spatial domain as the premise,and coprime sampling in the time domain.In the same time,considering that the sampling structures mentioned above are all based on identical temporal parameters,this thesis further proposes to allow different temporal parameters to be selected at different array elements,and the space-time sampling method proposed in this thesis is recorded as Spatial-Temporal Coprime sampling with Different temporal parameters(STCD).When the maximum achievable number of physical antennas in space and sampling points in time domain are limited,the degree of freedom is used as the cost function to optimize the space-time domain sparse parameters.In addition,considering that the DOA and Doppler frequency can be regarded as discrete parameters,and spatial smoothing will lose half of the virtual aperture,so we employ the compressed sensing algorithm to establish a joint estimation problem.Finally,compared with the spatial-temporal coprime sampling structure based on identical temporal parameters,the advantages of the STCD sampling method are illustrated in terms of degrees of freedom,resolution and overall performance.