In order to realize the elevation angle estimation for low-altitude targets at a low computational cost, a generalized multiple signal classification(GMUSIC) algorithm based on unitary transform is proposed, i.e., the DU-GMUSIC algorithm. Firstly, the covariance matrix of received data is used to construct the CentroHermitian matrix. Then, the real-domain GMUSIC algorithm is used to implement the initial angle estimation, and the multipath attenuation coefficient is calculated in succession. Finally, the attenuation coefficient is taken into account in the GMUSIC method to carry out the secondary angle estimation which is beneficial to further improvement of the angle estimation accuracy. This method can meet requirements of low-angle accuracy as well as lower computational burden. Simulation results prove the correctness and effectiveness of the proposed algorithm. Moreover, field experiment data are used to further validate the effectiveness of this method. In order to realize the elevation angle estimation for low-altitude targets at a low computational cost, a generalized multiple signal classification (GMUSIC) algorithm based on unitary transform is proposed, ie, the DU-GMUSIC algorithm. Firstly, the covariance matrix of received Finally, the real-domain GMUSIC algorithm is used to implement the initial angle estimation, and the multipath attenuation coefficient is calculated in succession. Finally, the attenuation coefficient is taken into account in the GMUSIC method to carry out the secondary angle estimation which is beneficial to further improvement of the angle estimation accuracy. This method can meet requirements of low-angle accuracy as well as lower computational burden. experiment data are used to further validate the effectiveness of this method.