This paper is mainly to deal with the problem of direction of arrival (DOA) estimations of coherent sources in impulsive noise fields modeled as α-stable distribution. Two new algorithms based on singular value decomposition (SVD) and fractional lower order statistics are proposed. The proposed methods rebuild a matrix by using the eigenvector corresponding to the largest eigenvalue obtained from the fractional lower order statistics matrix of the received data, then the auto-covariance matrix of the rebuilt matrix can be obtained. The noise subspace can be gained by SVD of the matrix which is reconstructed by the equal of the obtained auto-covariance and its complex-conjugation inversion. Simulation results demonstrate the effectiveness of the proposed methods for estimation DOA of the coherent sources in impulsive noise.