Fast DOA Estimation of Coherent Signals without Knowing the Number of Sources

A novel decorrelation algorithm that can eliminate the computation without having aperture loss is presented.Through vector reconstruction by eigenvector of maximum eigenvalue,the algorithm can deal with coherent signals.By utilizing high order power of the inverse matrix,the noise subspace can be approximate without knowing the number of signals.At the same time,the computation of algorithm is low because there is no eigendecomposition.Finally,the computer simulation confirms the validity of the proposed algorithm.