| Direction of Arrival(DOA)estimation is an important branch in the field of array signal processing.Uniform Circular Array(UCA)can provide information of both azimuth Angle and pitch Angle.With the same array number,the array aperture can be extended by virtualizing more array elements.In DOA tracking algorithm,the size of forgetting factor has a great influence on the tracking effect.Coprime Circular Array(CCA)retains the advantages of uniform circular array and Coprime circular array(CCA).In this paper,a two-dimensional DOA estimation and tracking algorithm for spatial information sources is studied based on the mutual prime circular array.The specific work is as follows:(1)The coprime circular array model is studied.Firstly,the array structure of the mutual prime circular array is introduced,the mathematical model of the mutual prime circular array is given,and the array prevalence matrix of the mutual prime circular array is derived.The twodimensional DOA estimation algorithm of the mutual prime circular array is studied,and the MUSIC algorithm of the mutual prime circular array is designed.The algorithm performs eigenvalue decomposition of the covariance matrix,and constructs the spectral peak search formula to search the two-dimensional spectral peak to estimate the incidence Angle of the signal.Compared with the uniform circular array,the DOA estimation performance of this algorithm is better.In order to avoid the search process of two-dimensional spectral peak and reduce the computational complexity,a two-dimensional DOA estimation algorithm based on the reduced dimension MUSIC algorithm in the circular array of mutual prime is studied.This kind of algorithm only needs one local spectral peak search,which reduces the algorithm complexity compared with two-dimensional MUSIC algorithm.Under the same array structure,the estimated DOA performance of this algorithm is close to that of the MUSIC algorithm.When the angles of multiple incident signals are similar,the algorithm still works effectively.The dimensional-reducing spectral peak search algorithm reduces the complexity of spectral peak search,but still requires a spectral peak search.In order to avoid the computational complexity caused by spectral peak search,the ESPRIT algorithm based on rotation invariance of two-dimensional signal parameter estimation in the mutual prime circular array is studied.This algorithm uses rotation invariance to estimate signal Angle without searching spectral peak,which further reduces the complexity of the algorithm.(2)In this paper,two-dimensional DOA tracking algorithm of mutual-prime circular array is studied.The DOA tracking algorithm based on subspace update of the circular array is studied.This kind of algorithm avoids the eigenvalue decomposition process in the process of classical higher resolution algorithm.The subspace of the circular array received signal is updated by the covariance update formula.The influence of the size of the forgetting factor on the tracking effect is studied.The forgetting factor determines the proportion of the historical and current sampled data in the process of updating the subspace,and thus affects the DOA tracking effect.Therefore,a time-varying forgetting factor is proposed,which can adjust the size of the forgetting factor adaptively according to the Angle change,and improve the tracking accuracy of DOA. |
