Research On Direction Of Arrival Estimation For Extended Sparse Array Design

Direction of arrival(DOA)estimation has long been viewed as an important research topic in contemporary array signal processing.The traditional DOA estimation algorithms are mostly based on the uniform linear array(ULA),leading to the resolved sources being less than the sensors.Hence the ULA is no longer satisfying the actual demand with the wide application of DOA estimation in various fields.Sparse array is a kind of structure that can effectively expand the array aperture to supply much more degrees of freedom(DOFs)than ULA,which provides the possibility of achieving DOA estimation that can resolve more number of sources and achieve higher precision.Therefore,this paper intends to further improve the performance of sparse array in DOA estimation through the design and optimization of array structure based on the classical sparse array structure.The main contents in this paper are divided into the following aspects:(1)A two-dimension multi-level dilated nested array which exploiting synthetic aperture technology,which aims to the three-dimension DOA estimation,is proposed.The proposed array structure applies 2D NA to the moving platform.The sensors position of the proposed structure is similar to the 2D NA while the interelement spacing between adjacent sensors of each row is expanded.After the phase correction operation,the signals received at multiple moments are combined for processing,thus the DOFs that can be obtained is further improved.The relationship between the multiple of interelement spacing expansion and the total number of moments for received signals is analyzed in details.(2)The NA with coprime relation(CP-NA)is proposed by giving full consideration of the virtual sensor positions provided by nested array(NA)and coprime array(CPA).According to the property of the fourth-order cumulants,the CP-NA is composed of two subarrays with coprime relationship,each subarray can be regarded as a NA with enlarged array spacing.By analyzing the distribution of the virtual array elements in the fourth-order difference coarray,the closed-form expression for the relationship between the number of sensors and the uniform DOFs of the proposed sparse array is derived.Experimental simulations are conducted to verify the effectiveness of the proposed two array structures in DOA estimation.Compared with the existing sparse array structures,more resolved sources and higher DOA estimation accuracy can be obtained by using the proposed two array structures.