Research And Implementation Of Nonuniform Array Direction Finding Technology

Subspace class DOA estimation algorithm is a very important method in array direction finding technology,because it has advantages of high resolution and high precision,it attracts more and more attention and research by many scholars.In the actual project,the noise subspace class algorithm represented by MUSIC algorithm has been studied and implemented by many scholars,however,the signal subspace class algorithm represented by ESPRIT algorithm is rarely used in project.The reason is that the complexity of the traditional MUSIC class algorithm based on the maximum likelihood estimation is lower than that of the ESPRIT algorithm.The ESPRIT algorithm is efficient but only applies to the invariant array.Due to element damage,channel damage and element position Offset and other reasons the uniform circular array is uneven array,causing that it can not be measured.Therefore,this paper carries out research focusing on the above problems,the main work is as follow:Firstly,the basic theory and algorithms of spatial spectrum estimation are introduced,including array signal processing principle,signal model,array model,power iterative algorithm and classical DOA estimation algorithm.Then,after the performance of non-uniform linear array and non-uniform circular array is simulated and analyzed,it finds that the spacing of non-uniform linear array elements is not limited by half-wavelength,and the aperture of the array can be expanded to a great extent.Non-uniform linear array is not only better in the array direction,angle resolution and direction accuracy than the uniform line,but also has anti-ambiguity.Even if the array is missing or the element position is shifted,there is no need to repair and correct the array,but just properly adjust the orientation vector.The direction finding performance near the same number as that of the uniform circular array.Secondly,based on focused research of low complexity and non-uniform linear array of ESPRIT algorithm,a non-uniform linear array of iteration ESPRIT direction finding scheme is put forward.The scheme uses the power iterative algorithm to estimate the signal subspace,and uses the phase compensation in the signal subspace of the nonuniform line array,and iterates to solve the direction of the DOA.Then the algorithm is simulated and analyzed by MATLAB.The iterative ESPRIT algorithm is not only comparable to the MUSIC algorithm in the direction finding accuracy,but alsothe algorithm is much simpler than the MUSIC algorithm in algorithm implementation.Finally,the algorithm of ESPRIT algorithm proposed in this paper is implemented in FPGA,and the relationship between accuracy,speed and resource is calculated.The main sub-variance matrix module,the power iterative algorithm module and the iterative solution module are divided into three parts.And the connection scheme is proposed for each module and the performance of the whole design is analyzed,including the accuracy of measurement,resource consumption,algorithm execution speed and error analysis.