Power and DOA Estimation with Newton-Like Method

To increase the capacity and link performance of mobile communications, it is important to accurately estimate the powers and directions of arrival (DOA) of sources. A multi-dimension nonlinear cost function is proposed for this purpose, where the DOAs and powers of sources are optimization parameters. The optimal results are obtained with the Davidson-Fletcher-Powell (DFP) algorithm and are compared with the results of alternating projection (AP) algorithm. The simulation results show that DFP has the accuracy in accordance with AP when sources are away from each other, but higher than AP when sources are near.