A new approach to parameter estimation for coherently distributed source

We consider the problem of estimating the parameters-the central direction of arrival (DOA) and angular spread of a coherently distributed source. This problem is encountered due to the presence of local scatterers in the vicinity of a transmitter or due to signals propagating through a random inhomogeneous medium. Since the computational complexity of the parameter estimation is normally highly demanding, we first decouple the central DOA from angular spread. And then the preliminary estimate of the central DOA can be obtained at a signal subspace. Angular spread is estimated by beamspace propagator method sequently. So a two-dimensional maximization problem is replaced by two one-dimension problem. The proposed algorithm has improved precision, low computational cost. Simulations clearly demonstrate that the algorithm is not only effective, but also enjoys better performance.