DOA estimation based on multi-resolution difference co-array perspective

This paper presents two kinds of K-level co-prime linear array geometries and the corresponding direction of arrival estimation algorithm based on the multi-resolution difference co-array (MRDCA) perspective. The MRDCA can simultaneously improve the degree of freedom and the angle-resolution by utilizing a class of virtual sparse uniform linear arrays generated by vectorizing the covariance matrix of the received observations of the K-level large scale sparse array. Compared to the prior two level co-prime/nested arrays, the aperture and the angle-resolution can be significantly increased with Kth power law for the K-level array, while the dimension of its scanning space is reduced to 1/K resulted from the spatial aliasing of the MRDCA. As a result, a low-complexity DOA estimation algorithm is proposed by combining a multi-resolution estimation at each level of sparse MRDCA and a followed probability decision strategy which aims at effectively identifying the genuine DOAs and excluding the replicas. In the end, the simulation results are provided to numerically validate the performance of the proposed array geometries.