In mobile communication systems, the number of sources, source types, direction-of-arrival (DOA) and power parameters are important for the downlink precoding, power control and user identification. In this letter, a multi-parameters estimation approach is introduced for coherently distributed (CD) sources under the scenario of mixed circular and noncircular signals. Being different from the traditional subspace based approaches, the proposed approach first constructs a perturbed ${\ell _{2}}{\mathrm{ - }}{\ell _{F}}{\mathrm{ - }}{\ell _{0}}$ norm minimization problem in sparse representation framework, and then solve it by jointly exploiting truncated ${\ell _{1}}$ function approximation, cost function decomposition and iterative strategy. Finally, multiple parameters are estimated by the output sparse vectors and sparse perturbed matrices. Simulations validate the effectiveness of the proposed approach.