Research On Doa Estimation Based On Coprime Array

Direction of arrival(DOA)estimation has always been an important technology in array signal processing and is widely used in radar,navigation,and communications.Nowadays,most of the DOA estimation algorithms are proposed on the basis of uniform linear array(ULA).The theory is widely and deeply researched,however,it is limited by the problem of the spacing and mutual coupling of the array elements.As a hot research topic in recent years,the coprime array has the advantages of high degree of freedom and less array elements,and breaks through the limit of the estimated number of sources less than the number of elements.It can also estimate more sources on the basis of improving the accuracy and resolution of the incidence angle estimation.The time-frequency analysis and processing method can distinguish signals on the time-frequency plane,which has advantages in the processing of non-stationary signals.Therefore,this paper will use the coprime array as a model,and then combine the time-frequency analysis method with the Direction of Arrival(DOA)estimation method to explore more accurate methods of DOA estimation based on the coprime array.In this paper,several traditional wave direction estimation techniques based on uniform linear array are introduced.The conventional beamforming(CBF),Capon algorithm and multiple signal classification(MUSIC)algorithm are introduced on the premise of non-coherent sources,and the MATLAB simulations and performance analysis are carried out respectively.It also introduces several decoherent algorithms under coherent source conditions,including forward and backward smoothing methods,improved MUSIC algorithms,etc.Finally,their performance is analyzed and compared.The following is the introduction of the model of the coprime array and the concept of the position difference and degree of freedom(DOF)for the coprime array.The structure of the array can be divided into the original structure and the extended structure.Two traditional algorithms of MUSIC and DECOM are introduced in detail for the estimation of the direction of arrival of the original structure.Through performance analysis,the traditional algorithm based on the original mutual qualitative array can not make full use of the degree of freedom,and the scope of application is narrow.In order to increase the number of degrees of freedom as much as possible to increase the number of sources that can be estimated,a coprime array of extended structures is introduced.However,due to the change of the structure of the data covariance matrix,the traditional algorithms are no longer applicable.To this end,several improved algorithms have been introduced.These methods all require correcting the covariance of the data and then performing subsequent DOA estimation steps.And the above algorithm is simulated and verified by MATLAB analysis.In order to further improve the accuracy of source estimation,the time-frequency information is combined with the DOA estimation of the coprime array.By replacing the data covariance matrix with the time-frequency distribution matrix obtained by the bilinear time-frequency transform and combining with the improved coprime array algorithm,a time-frequency smoothing MUSIC method based on a coprime array can be obtained.In this way,the signals can be distinguished on the time-frequency plane and processed separately to further improve the angular resolution and estimation accuracy.