Research On Signal Source Estimation And Direction Finding Algorithm Based On Uniform Circular Array

The estimation of Direction of Arrival has a wide range and important backgrounds in the fields of radar,communication,detection and medicine.It is a key problem in parameter estimation of array signal processing.The most commonly used planar array layouts in the array direction finding are uniform linear array and uniform circular array.Compared with the linear array,the uniform circular array has the advantages of small aperture,uniform measurement accuracy and can be used to two-dimensional direction finding.In this paper,we use the uniform circular array as the array model to study the source number estimation algorithm,the direction finding algorithm under the spatial spectrum estimation and the time difference direction finding algorithm.In the current spatial spectrum estimation theory,many direction finding algorithms need to know the number of signal sources.In this paper,we study the method of estimating the number of sources based on information criterion.For this method cannot obtain the accurate estimation of the number of sources under the color noise environment,we study the diagonal loading algorithm based on covariance matrix.On the premise of correct estimation of the number of sources,the classical MUSIC direction finding algorithm is also studied.Although the MUSIC algorithm can obtain the high precision direction finding value of the multi-source source,it is not suitable for the coherent source.The commonly used uniform circular array coherent source direction finding method needs to be transformed into the virtual linear array,and then the spatial smoothing algorithm is used to solve signal coherence.This kind of method is complex and sacrifices the aperture of the array.Therefore,the virtual array translation method is studied and simulated in the paper.The algorithm can realize direction finding without sacrificing the aperture of the array in the coherent environment.The signal model of spatial spectrum estimation is based on the narrowband source.For the direction finding of wideband signal,the classical algorithm includes Incoherent Signal-Subspace Method and Coherent Signal-Subspace Method.These two algorithms convert the broadband signal model into a narrowband signal model,and then use the spatial spectrum mapping algorithm for direction finding.These algorithms are complex and computationally intensive.Therefore,this paper studies the direction finding problem of broadband signal from the perspective of time difference algorithm.When the uniform circular array is applied to the platform on board or on plane,the delay difference between the elements is very small,which may be the same order of magnitude as the sampling interval.In order to obtain the high precision time difference estimation without increasing the number of sampling points,this paper applies the fractional delay estimation method to improve the accuracy of the non-integer part of the sampling interval of the delay difference,thus realizing the high precision direction finding of the wideband signal in the small aperture circular array.In the process of signal processing,the number of sampling points must satisfy the Nyquist sampling law.But the sampling of high frequency wideband signal requires too many points,which increases the system storage burden and resource waste.Aiming at this problem,this paper proposes a new broadband signal time difference measurement algorithm based on Compressed Sensing theory,which can obtain the direction finding accuracy similar to fractional delay estimation algorithm with a small number of observation points.The proposed algorithm can save the storage space and greatly reduce the algorithm execution time and improve the efficiency of the algorithm at the same time.