An new algorithm for DOA estimation based on non-orthogonal joint diagonalization

In this paper, A set of spatio-temporal covariance matrices which possess diagonal structure are introduced. Then we use a high quality signal subspace as a dimension-reduction matrix dealing with the set of correlation matrices. A novel cost function was proposed based on nonlinear least squares. Afterwards, a new approach termed Tri-iterative algorithm (TIA) was derived for solving the cost function and estimating the direction of arrival (DOA) of sources. Some numerical simulation examples are provided to demonstrate the effectiveness of the developed algorithm.