Abstract: This paper presents a new Direction-Of-Arrival (DOA) and power estimation algorithm based on two-stage weighted L1-norm penalty. In this scheme, this paper first obtains a better statistical performance vector measurement model by sum-average arithmetic using array covariance matrix elements. Then the sparse representation of the new model corresponding to the related overcomplete basis is successively constructed. This paper exploits MUSIC spectral function and initial estimation result as the weights to construct two-stage weighted L1-norm penalty reconstruction method. Consequently, this paper achieves the DOA and power estimation of multiple sources. Moreover, a reasonable regularization parameter is also selected by2 distribution function. The simulation results demonstrate that the proposed algorithm is able to provide higher resolution and estimation accuracy than the other state-of-the-art methods.