The Study Of Direction Of Arrival Estimation Algorithm Based On Subspace Decomposition

Direction of arrival(DOA)estimation technology originates from the development of spatial filtering and time domain spectrum estimation.Its performance in parameter estimation is superior,therefore,whose application prospect is broad,such as radar,sonar,electronic countermeasures and wireless positioning technologies.It has currently become one of the important research fields of array signal processing.The subspace decomposition algorithms,in many kinds of DOA estimation algorithms,estimate parameters based on the output covariance matrix eigenvalue decomposition.Due to the algorithm can achieve high estimation precision and resolution;it is valued by many scholars.MUSIC(Multiple Signal Classification)algorithm implements flourish of the super resolution direction-finding technology.It promotes the development of subspace decomposition algorithm.ESPRIT(Estimation of Signal Parameters via Rotational Invariance Techniques)algorithm is put forward,which is another significant progress in subspace decomposition algorithm.ESPRIT algorithm is mainly using the feature of signal subspace rotation invariance in DOA estimation,compared to MUSIC algorithm,which has a small amount of calculation and without spectral peak searching.The above two algorithms are representative algorithms in subspace decomposition.The former is based on noise subspace,while the later is based on property of rotation invariant subspace.These two kinds of algorithm performance are superior compared with early DOA estimation algorithm,and they have excellent estimation precision and spatial resolution,many scholars has carried on related technology research,improved algorithms of two algorithms emerge endlessly.In this paper,the DOA estimate current algorithm based on subspace decomposition was studied,and the characteristic of algorithms were analyzed.Those algorithms are improved on the estimation precision and the amount of calculation.Thus two improved algorithms are presented in this paper.Common arrays structures are studied and special arrays structure is designed to be suitable for improved algorithms.The main research work in this paper is as follows:(1)In this paper,the root minimum norm algorithm is improved and the root minimum norm algorithm based on real polynomial is proposed.The main idea of the improved algorithm is conformal mapping technology,which converts the complex polynomial to real polynomial,so as to reduce the amount of calculation and slightly improve the estimation precision.(2)In this paper,2D-Unitary ESPRIT algorithm is applied to the special array,whose structure has multiple sets of sub-arrays with translation invariant.Therefore it can calculate multiply candidate estimations.The DOAs are selected among the estimations based on the selection criterion with 2D-MUSIC spectrum,which improves the estimation precision.(3)ESPRIT algorithm is only applied to array structure with translation invariance,and hexagonal array possess multiple combinations of sub-arrays.The optimal sub-arrays are used to estimate signals in different directions,which ensure the final DOAs are optimum.But array elements are arranged crowd and the number of elements in sub-array is less.In this paper,hexagonal star array structure is designed to improve performance of hexagram array.It can effectively use most elements in array,which can obtain better estimation results.Therefore,the estimation performance of hexagram star array structure is better than hexagonal array structure on circumstance of nearly elements number.