Skip to Main Content
This paper investigates a maximum likelihood (ML) approach for direction of arrival (DOA) estimation based on the algorithm suggested by the 3GPP for scalar channels in UTRA-TDD mode. The performance of the algorithm is assessed by resorting to simulations in typical UMTS scenarios. The results indicate that the proposed scheme overcomes the limitation of the angular resolution inherent to classical techniques like subspace or beam-forming methods. In particular, it is shown that waves exhibiting an arbitrarily small difference in azimuth and temporal spacing higher than one chip time interval can be easily separated. A threshold operation is included to select only the most significant paths, in terms of energy, of the estimated channel. The results show that a good performance can be achieved from low to high values of Eb/N0.