2-D DOA estimating of multipath signals by exploitation of cyclostationarity

In most wireless communication systems, two-dimensional directions of arrival (DOA) of multipath signals need to be found for spatial selective transmission. In this paper, we propose a new algorithm to estimate 2-D DOA of multiple narrow-band signals. A DOA cyclic matrix is constructed whose eigenvalues and eigenvectors can be simultaneously used to extract 2-D DOA without 2-D searches. By exploiting the temporal property of cyclostationarity, the signal detection capability is significantly improves. Besides, based on the decorrelation model for mobile terminal signals, the algorithm can be effectively extended to the coherent case without spatial smoothing and the loss of array aperture. Simulation results are given to illustrate the performance of the new algorithm.