A spatio-temporal stacking approach for estimating two-dimensional direction-of-arrival

A number of 2-D DOA estimation techniques based on L-shaped array have received much attention. In general, these methods require division of the L-shaped array into two independent uniform linear array (ULA) to obtain the azimuth and elevation angles independently, and then use additional pair matching technique to achieve 2-D DOA estimation. Therefore, these methods have some drawbacks such as 1) requirement of pair matching which may increase the computational burden significantly and 2) estimating the azimuth and elevation angles by independently using each ULA half of the L-shaped array. The purpose of this paper is to develop a spatio-temporal stacking approach (STSA) to deal with these shortcomings. The STSA method first partitions the many lag cross-correlation matrices into a lot of submatrices based on the assumption on the second-order temporal structure (SOTS) of the source signals, and then stacks these submatrices according to some special structure to form a spatio-temporal stacking matrix. Finally, the joint singular value decomposition (JSVD) technique is exploited to extract out the 2-D DOAs one by one.