Research On DOA Technology Of Small Aperture Array Based On Compressed Sensing

In recent years,the fields of military,civil and modern medical treatment have all involved array signal processing.Its purpose is to process echo signals through different algorithms,suppress useless interference and noise,highlight useful signals,and obtain people's senses from them.Information of interest.The field of spatial spectral evaluation,namely DOA technology of array signal processing,is widely used in practical applications,mainly for accurate estimation of signal parameters accepted in space.In contrast to a large aperture array having an aperture size of several tens of wavelengths,a small aperture array refers to an array whose array aperture is equal to or only a few wavelengths of the signal wavelength.At this time,when the signal DOA is estimated,the estimation performance of the algorithm will be deteriorated.Compression Sensing(CS)does not require a sampling rate that doubles the signal bandwidth,it only needs a low sampling rate to achieve signal reconstruction,which is consistent with the properties of small aperture array.In this paper,the DOA estimation problem in the background of small aperture arrays is analyzed in depth for the constraints of small aperture arrays.The CS-based DOA estimation algorithm and its meshing problem are studied.The two backgrounds are incoherent signals and coherent signals.Simulations have been carried out under the following conditions.The results show that CS-based DOA estimation can overcome the adverse effects caused by small apertures,improve resolution and estimation accuracy,has strong adaptability and has important application value.On this basis,the realization method of CS-based small aperture array DOA estimation is studied.The part of the algorithm that is suitable for parallel computing during the operation of the algorithm is implemented with the GPU,which reduces the calculation time and speeds up the operation of the algorithm.First of all,this article modeled and analyzed the array signal model of the small aperture array,and used the MUSIC algorithm to prove the constraint of the small aperture array on the DOA estimation.This paper briefly explained the principle of compressed sensing,its theoretical framework and its applicability in solving DOA evaluation problems.The logarithmic summation algorithm and OMP algorithm are used to estimate the DOA of the small aperture array.On this basis,an improved logarithm summation algorithm is proposed to reduce the program running time.Aiming at the problem of grid mismatch,this paper comparatively analyzes three grid division methods and selects the one with the best effect.In order to perform DOA evaluation on the background of non-coherent and coherent signals,this paper analyzes the performance of the algorithm by comparing and analyzing the simulation results of several classic algorithms and CS algorithms.forthe problem of false peaks when the classic algorithm estimates coherent signals,three performance improvement schemes were proposed: smooth MUSIC algorithm,higher-order spectral method and the improved logarithmic summation algorithm,which are analyzed and simulated.The experimental results prove that the three schemes all have better improvement effects.Finally,this paper studies the implementation method of DOA estimation of small aperture array based on CS.Some of the algorithms suitable for parallel computing are applied to the GPU,and compared to the CPU calculation simulation results.The simulation results show that the results of the two calculations are exactly the same,but the GPU calculation time is shorter.