Study On DOA Estimation Algorithms With Improved Nested Array

With the development of DOA estimation theory,non-uniform arrays not only increase the degree of freedom and improve the Angle resolution,but also realize the underdetermined DOA estimation that the number of sensors is less than the number of sources under the same number of array elements.Therefore,it has become a hot topic in array study.In view of the current structure of nested array of possible problems in engineering application,A kind of array elements position relatively flexible non-uniform array is proposed,and on the basis of the array to realize the space sources of 1-D and 2-D DOA estimation,the final will be more than its application to the MIMO radar parameters estimation,can achieve more accurate positioning to the sources.The specific works are summarized as below:1.Through the geometric operation of the number of array elements,the two-level nested array can achieve the maximum degree of freedom(the maximum number of virtual array elements)of the array.Because the two-level array are composed of uniform linear arrays and the first-level array elements are closely arranged,it is easy to cause the problems of engineering installation and mutual coupling.Aiming at this problem,A non-uniform array antenna structure is proposed,the array don't need to be composed of two or more uniform linear arrays,so there is no close interval array elements,can effectively reduce the mutual coupling between array elements,On the premise of satisfying the same performance as nested array(the same number of consecutive virtual array elements as nested array),compared with the nested array,this method is more flexible,more widely engineering application.2.For the proposed non-uniform array,an algorithm for processing redundant data sets is studied.The final output parameter of this algorithm is a virtual array receiving data without redundant elements information,and the data is provided to the subsequent DOA estimation algorithm for processing.3.In order to avoid the loss of degree of freedom caused by spatial smoothing algorithm,sparse reconstruction algorithm is combined with 1-D DOA estimation under non-uniform array,which can not only realize underdetermined DOA estimation,but also have betterestimation performance under the condition of low SNR and snapshot number.But in the practical engineering application,want to achieve more accurate positioning,it is not enough to know target of elevation angle information,therefore,we are on the basis of 1-D DOA estimation,the extension of the non-uniform linear array to L-shaped array,this paper studies 2-D DOA estimation algorithm without eigen decomposition.In the case of large number of snapshots and many array elements,the algorithm can effectively reduce the processing time.The simulation results show that the 2-D DOA estimation under L-shaped non-uniform array can not only achieve high estimation accuracy but also deal with close range targets.4.In order to further expand the MIMO radar array of degrees of freedom,will the proposed array is configured to transmit and receive array,this paper introduces the bistatic MIMO radar signal model under non-uniform array,and according to the signal model of a sparse reconstruction method is introduced,through mining the array manifold matrix relation of transmitting angle and receiving angle,a high-dimensional dictionary to two lower dimensional dictionary,not only reduces the computation of sparse reconstruction algorithm,achieved from high-dimensional sparse signal to low dimensional mapping,and through the lower dimensional sparse signal reconstruction completed angle pairs automatically.The simulation results show that the proposed method can achieve joint DOD and DOA estimation of multiple targets in a single snapshot,and can also process spatial coherent target sources.